Breaking New Ground: Space Agencies and Epidemiologists Partner Up on Particulates

Abstract

When cold air traps pollutants over Johannesburg on winter afternoons, a brown haze can form, with particulate matter levels that exceed South Africa’s ambient air quality standards.¹ In mid-June 2022 (early winter in South Africa), an inversion also magnified a foul smell that left one resident tweeting, “Johannesburg is a toxic cesspool. It now smells like rotten eggs.”² Although the gas responsible for that smell, hydrogen sulfide, is not regulated in South Africa, its noxious odor was a reminder that levels of other pollutants might also be high.³

Smog blankets Johannesburg on a sunny day in 2019. The capital city is located in South Africa’s Highveld—a high-altitude inland area that is the country’s industrial heartland.³ Image: © iStock.com/Rich Townsend.

It is hard to capture the spatial and temporal variations in air pollution at specific spots in Johannesburg. The region’s sparse monitoring system measures the total mass of particulate matter in the air but not the composition or size of specific particles. “We really don’t know exactly what’s in it,” says Kristy Langerman, an atmospheric scientist at the University of Johannesburg. “But I can speculate and say it’s from traffic as well as some domestic burning and dust.”

In the coming years, Langerman and other air pollution experts may get some answers, thanks to an upcoming U.S. National Aeronautics and Space Administration (NASA) satellite mission, which has become a collaboration with the Italian Space Agency (ASI). The mission will deploy the Multi-Angle Imager for Aerosols (MAIA) instrument to capture information about the size and composition of particulates in the atmosphere. MAIA is anticipated to launch in late 2024, according to sources interviewed for this story.

Nothing Straightforward about Data Collection

One of those sources, David Diner, is a senior research scientist at NASA’s Jet Propulsion Laboratory and principal investigator on the MAIA project. As NASA’s first competitively selected Earth-observing mission to include epidemiologists on the scientific team, Diner says the project will help support studies of health effects from short-term, long-term, and gestational exposures to various air pollutants.

For some localities, the information MAIA will provide could be revolutionary. Studies are showing that fine particulate matter (${\mathrm{PM}}_{2.5}$) may cause detrimental health effects at levels lower than originally thought.^4,5 The World Health Organization (WHO) has dropped its annual average ${\mathrm{PM}}_{2.5}$ exposure guideline from $10{\mathrm{\mu }\mathrm{g}/\mathrm{m}}^3$ to $5{\mathrm{\mu }\mathrm{g}/\mathrm{m}}^3$.⁶ In the United States, the Biden administration convened a scientific panel in 2022 to discuss lowering the U.S. annual average exposure standard for ${\mathrm{PM}}_{2.5}$ from $12{\mathrm{\mu }\mathrm{g}/\mathrm{m}}^3$ to as low as $8{\mathrm{\mu }\mathrm{g}/\mathrm{m}}^3$.⁷

According to several epidemiologists working with MAIA—such as Beate Ritz from the University of California, Los Angeles—public health officials in low- and middle-income countries (LMICs) might not have as much influence as those who advise the U.S. president or the WHO. “Many government leaders will say, ‘What you see in the West does not apply to us; our situation is different,’” says Ritz. “But with MAIA’s data, hopefully, health researchers in those countries will be able to finally pressure their governments into action.”

Pros and Cons of Satellite Data

More than two decades ago, Diner helped design the Multi-angle Imaging SpectroRadiometer (MISR), which monitors how aerosols and clouds affect climate.⁸ MISR was designed to view Earth from nine fixed angles and in four spectral bands (blue, green, red, and near-infrared), with a spatial resolution of a few hundred meters. Its data offer a detailed picture of how the abundance of atmospheric particulate matter varies in space and time, among other outputs.⁹ MAIA will add to that by using a specialized digital camera that can be pointed at selected view angles and record light in 14 spectral bands.

MISR images over India in 2016. Left, the vertical viewing camera captured the Himalayas, which tend to concentrate pollution to the south. New Delhi was under such thick haze that aerosol optical depths, right, were not calculated because the algorithm classified the haze as a cloud.³³ Image: Courtesy of MISR Team, NASA/GSFC/LaRC/JPL.

Diner explains that the additional spectral bands will yield greater insight into particle composition. For instance, the shortest (ultraviolet) bands are sensitive to particles containing certain types of minerals and organic matter; the visible and near-infrared spectra are primarily sensitive to fine particles. Shortwave infrared bands will provide information about coarser particles, such as dust grains and volcanic ash. Additional bands measure polarization, providing another means of understanding particle properties.

“We’re basically looking at scattered sunlight and using the characteristics of that sunlight—its variation with wavelength and angle—to infer the physical and optical properties of the particles that are scattering the light,” says Diner. The shape of particles—whether spherical, like liquid droplets, or irregular, like dust grains—affects how they scatter sunlight, he explains. “All of these factors together make this multi-angle method, in conjunction with the multispectral and polarimetric measurements, pretty powerful for helping to identify specific pollutants.”

Nevertheless, it is tricky to translate satellite data into accurate estimates of what humans breathe at ground level. “Satellite-based ${\mathrm{PM}}_{2.5}$ data are not direct measurements,” explains Meredith Fowlie, an energy economist at the University of California, Berkeley. “They are noisy estimates, and it requires a lot of scientific modeling, assumptions, and statistical modeling to construct the [ground-level] PM measures.”

Joel D. Kaufman, a professor of environmental health at the University of Washington and editor-in-chief of Environmental Health Perspectives, takes it a step further. “Even when the satellite estimates can get pollution concentrations correct at a $1\times 1\mathrm{km}$ resolution—which is much better than no data at all—they are missing the finer scale of important pollutant gradients that exist over tens of meters from important sources,” he says. “Sources like roadways and waste incineration can be important at this scale.”

MAIA will acquire multiple-angle imagery of selected target areas both along the orbit path as well as across it. Image: Courtesy of NASA/Jet Propulsion Laboratory, California Institute of Technology.

Diner and the MAIA team acknowledge the trade-offs between satellite and surface-based approaches. “With certain pollutants, the spatial variability is quite significant, and so a straight interpolation could, itself, lead to substantial errors,” he says. But there are simply not enough ground-based monitors in many parts of the world—especially in LMICs—to determine what residents are exposed to on a daily basis, he explains.

Target Areas

The mission will target 11 primary areas—regions surrounding Los Angeles, Atlanta, Boston, Barcelona, Rome/Bologna, Addis Ababa, Johannesburg/Pretoria, Tel Aviv/Haifa, Delhi, Beijing, and Taipei/Kaohsiung—and approximately two dozen secondary sites.¹⁰ The group will use data from both satellite- and ground-based instruments to generate pollution maps. But not every target city has a network of speciation monitors for determining chemical species and amounts of collected particles.

MAIA researchers will use the speciation monitors already in place in the primary target areas in the United States, Spain, and Italy, Diner says. And thanks to a grassroots program called the Surface Particulate Matter Network (SPARTAN), which works to put air pollution monitors in international locations, three other cities (Pretoria, Tel Aviv, and Beijing) received speciation monitors before the MAIA project started.¹¹ Additional monitors are now sited in Johannesburg, Addis Ababa, Haifa, Delhi, Taipei, and Kaohsiung. The MAIA team is putting aerosol mass and optical depth monitors (another type of filter-based sampler) in several other primary target cities.

“The result will be that each primary target area will have at least two speciation monitors capable of collecting particulate matter on filters, which are then analyzed in the lab for sulfates, nitrates, elemental carbon, organic carbon, and dust,” Diner says. Aethalometers, which measure airborne black carbon concentrations in real time, have been placed in Addis Ababa, Delhi, and Beijing. Such ground-based measurements are essential to MAIA’s approach for converting the satellite measurements into reliable estimates of pollution concentrations. The U.S. Department of State has assisted with the shipment and operation of many of these sensors, and the U.S. Agency for International Development is supporting the ground-based measurements and chemical analyses in MAIA’s primary target areas in Africa.

The MAIA team is supporting other creative solutions to the scarcity of monitors. For instance, Addis Ababa is the primary target area with the fewest air quality monitors, so MAIA researchers have installed a network of PurpleAir monitors—low-cost sensors that measure total ${\mathrm{PM}}_{2.5}$ and provide citizen scientist and community groups with measures of local air quality.¹²

Amid visible smog, children play in the street in Addis Ababa on 3 February 2019, a car-free day promoted to help reduce air pollution. MAIA will bring new ground-based monitors to the city to complement satellite data. Image: © Eduardo Soteras/AFP via Getty Images.

Chemical transport modeling for each target area will factor in emissions from both anthropogenic and natural sources, the temporal and spatial distribution of those emissions, and their chemical reactions. Such modeling can account for wind movement, trapping by cold air, and other local weather phenomena.¹³ The resulting estimates will help fill in spatial and temporal gaps in the satellite coverage.¹⁴ “It’s really three points on the triangle,” Diner says. “The satellite, the surface monitors, and the model [are] all working together.”

The newly minted NASA-ASI collaboration means that more of Italy will potentially be studied than had been proposed in the original MAIA mission. Massimo Stafoggia, an epidemiologist with the Lazio Regional Health Service in Rome and long-time MAIA team member, says the central part of the country, including Rome and the Po Valley—which, due to traffic, industry, and geography, has the highest levels of ${\mathrm{PM}}_{2.5}$ in all of Europe—will remain one of the mission’s target areas. But, he says, ASI researchers now hope to study some secondary target areas in the north and in the south—including Taranto, home to one of Europe’s largest steel factories.

“There is a large population living very, very close to that plant,” Stafoggia says, “and because of that, it’s a very hot topic in that area.” He is involved in several longitudinal population-based air pollution studies, including some where investigators have tried to estimate the components of ${\mathrm{PM}}_{2.5}$ that residents are breathing in certain cities.¹⁵ “But,” he says, “I am pretty sure that the data made available [from] MAIA will be much more accurate.”

Collaborating with Epidemiologists

Diner says he has learned a great deal from working with the team’s epidemiologists to design the mission so its data will best support research on health effects. “When we were choosing our target areas, I would have thought to look only at the most polluted places,” he says. “But the epidemiologists said, ‘No, we want to look at clean places, too, because when you look at a $10\text{-}{\mathrm{\mu }\mathrm{g}/\mathrm{m}}^3$ change in an aerosol or particulate matter concentration where it’s relatively clean, it’s a much bigger change than in a place that’s very polluted.’ So they want to see the whole range.” Diner says MAIA will fly over each area at least three times per week.

MAIA team members who specialize in remote sensing helped Diner select the spectral bands. The team epidemiologists weighed in on other constraints. For instance, if there are two targets that could be observed in one area as the satellite passes overhead, which one would they prioritize?

Yang Liu, who studies air pollution at Emory University’s Rollins School of Public Health, leads the group that is integrating instrument data with atmospheric chemistry and spatial statistical models. The operational algorithm they create will translate the satellite signals into concentrations of the same pollutants collected by the speciation monitors, at a resolution of $1{\mathrm{km}}^2$.^16,17

Epidemiologists will compare those maps with local records to study health effects associated with ${\mathrm{PM}}_{2.5}$ exposure in the short and long terms, as well as during pregnancy.¹⁸ An estimated $4.1\text{\hspace{0.17em}million}$ deaths globally are attributed to ${\mathrm{PM}}_{2.5}$,¹⁹ and investigators have shown links between exposure to fine particles and “everything from simple asthma exacerbation²⁰ to heart attacks²¹ to cancer,²²” says Liu. “In recent years, it has been linked to cognitive changes²³ and birth defects.²⁴ It has such a broad spectrum of impacts.”

Birth outcomes are among the health effects that MAIA researchers will study. Preeclampsia and low birthweight are some of the pregnancy and childbirth concerns already linked to air pollution.²⁷ Image: © iStock.com/FatCamera.

Short-term exposures (daily to monthly) are perhaps the easiest to study. Liu says investigators in the Atlanta region will use MAIA data to generate pollution estimates, then compare those to emergency department visits for problems such as asthma attacks and cardiovascular disease. Long-term (multiyear) exposures will be compared with hospital or physician office visits for cardiovascular disease. Pregnancy-related exposures will be studied by comparing hospital or physician office visits for preeclampsia, low-birthweight infants, and other pregnancy and childbirth concerns with ${\mathrm{PM}}_{2.5}$ levels in the preceding 12 months. Given that adverse cardiovascular, respiratory, and birth outcomes have already been linked to air pollution,^25–27 these studies can confirm whether MAIA findings regarding ${\mathrm{PM}}_{2.5}$ constituents are indeed valid for health effects studies in places without ground-based monitors.

Electronic medical records simplify health studies, and Liu will have little trouble getting digital emergency department records in Georgia. However, not all hospitals in LMICs are set up to keep digital records.²⁸ In South Africa, for instance, public hospitals have struggled to adopt them,²⁹ so MAIA researcher Janine Wichmann, an epidemiologist at the University of Pretoria, will use records from private hospitals. This will skew the cohort toward more affluent patients, she says, “but at least, perhaps, one can then compare it to, say, European or U.S. studies.”

For pregnancy studies, MAIA researchers will need access to birth certificates. Ritz says MAIA-affiliated researchers in countries without electronic recordkeeping, such as India and Ethiopia, will have to gather paper birth certificates. “You have to have special outreach efforts,” she says.

Keeping It Local

Wichmann and her colleagues are pleased that the MAIA team has made partnering with local researchers a priority. “Parachute science is my pet peeve,” says Rebecca Garland, an atmospheric scientist at the University of Pretoria, referring to studies conducted by outside researchers without involving local expertise or sharing findings. “Parachute scientists don’t engage with local researchers or stakeholders, so their work, and thus their results, aren’t always relevant or needed. Even if it is cool science, if it is not a priority of the local stakeholders it won’t have any impact. As these [parachute] researchers are working alone, they don’t know the actual gaps or priorities.”

Once the MAIA data are available, Garland, Langerman, and Wichmann hope to see a more complete picture of how different pollutants move through the province at different times of the year. This will be important for establishing policies that protect health, Garland says. Many countries around the world have no air quality standards at all.³⁰ South Africa adopted air quality limits for eight priority pollutants in 2004 and began setting up a network of air monitors.³¹ That was a good start, says Wichmann, but more still needs to be done—South Africa’s air quality standards are less stringent than WHO guidelines. For instance, the WHO guideline for sulfur dioxide is an average of $40{\mathrm{\mu }\mathrm{g}/\mathrm{m}}^3$ over 24 hours, compared with the South African standard of $125{\mathrm{\mu }\mathrm{g}/\mathrm{m}}^3$. “On the vast majority of days, the WHO standard is being exceeded [in the province],” says Wichmann. “But according to South African officials, there’s no pollution problem.”

SPARTAN monitors are improving ground-based air monitor coverage in the MAIA primary target areas. Here, Siyabonga Simelane, a PhD student at the University of Johannesburg, stands with a new monitor installed by SPARTAN in the capital city (visible in the background). Image: Courtesy Kristy Langerman/University of Johannesburg.

Wichmann is dubious that any amount of data will trigger South African regulatory agencies to rethink the nation’s air quality standards. The country’s economy is struggling, she says, and politicians tend to think that the economy should take precedence over environmental standards.

Another epidemiologist on the MAIA project, Bart Ostro, from the University of California, Davis, offers a more optimistic view. He has seen officials change their outlook on air pollution once they are presented with analysis of local data. For example, in the 1990s, Ostro helped health researchers in Chile and Thailand determine that the particulate matter they were measuring was associated with a multitude of health effects.³²

“I was told later that [this research] had had a very important role in awakening scientists and politicians to the air pollution issue in these countries,” Ostro says. “Conducting studies in people’s own countries can really be a wake-up call.”

Biography

Nancy Averett writes about science and the environment from Cincinnati, OH. Her work has been published in Scientific American, Discover, Audubon, Sierra, and a variety of other publications.