Doctors and public health experts agree that breathing fine particulate matter (PM2.5) can be harmful to human health. The airborne particles—thirty times smaller than the width of human hair—can pass easily into the lungs and bloodstream, where they can increase a person’s risk of dying from heart disease, stroke, lung cancer, chronic obstructive pulmonary disease, and lower respiratory infections.
However, current estimates of the total number of premature deaths linked to PM2.5 range widely, from 3 to 9 million people each year. And there has long been uncertainty about the proportion of these deaths that are due to naturally occurring windblown dust versus human-caused (or anthropogenic) pollution, which comes from factories, transportation, power plants, cookstoves, crop fires, and other sources.
Research led by a team of atmospheric scientists based at NASA’s Goddard Space Flight Center indicates that the health burden associated with PM2.5 is somewhat lower than previous estimates suggest—and sheds light on the role of dust. The researchers—including Hongbin Yu and Alexander Yang—calculated the global health effects of PM2.5 by analyzing exposure over an extended period of time using a NASA atmospheric modeling system integrated with medical data from the Univeristy of Washington’s Global Burden of Disease Study.
The NASA team’s conclusion: exposure to PM2.5 likely contributed to 2.89 million premature deaths in 2019—1.19 million from heart disease, 1.01 million from stroke, 287,000 from COPD, 230,000 from lower respiratory infection, and 166,000 from lung cancer. According to their estimates, roughly 43 percent of those deaths occurred in China and 23 percent in India—two of the most populous and polluted countries in the world. Other countries with significant exposure to PM2.5 and large numbers of premature deaths included Pakistan, Bangladesh, and Nigeria—though none of these countries accounted for more than three percent of the total deaths linked to PM2.5.
The analysis linked 22 percent of the premature deaths associated with PM2.5 to dust—much of this in a “dust belt” that spans from West Africa to East Asia. “In both northern China and northern India, you have huge urban populations living downwind of major dust sources,” explained Yu. “You also have this in West Africa and the Middle East to some degree, especially in Nigeria and Egypt.”
The satellite image above shows a wall of dust from the Gobi Desert approaching northeastern China and the Beijing metropolitan area on March 10, 2023. The image below shows dust from the Thar Desert blowing east over the densely populated Indo-Gangetic Plain and mixing with smoke and haze from crop fires and urban pollution on April 7, 2021. Both images were acquired by the Moderate Resolution Imaging Spectroradiometer (MODIS) on NASA’s Terra satellite.
“This study is a reminder that dust—something that is largely natural and something that we can’t easily control with policy—can have an important impact,” said Yu. “In some countries in the dust belt, dust alone can push a population’s PM2.5 exposure well above World Health Organization guidelines.”
The team reached their conclusions by first calculating how much background exposure people in different parts of the world had to PM2.5 in 2019 by using a meteorological and atmospheric reanalysis system called the Modern-Era Retrospective analysis for Research and Applications, Version 2 (MERRA-2). MERRA-2 is a model that uses real-world observations to help simulate how dust and other key aerosol particles move and change in the atmosphere over time. The researchers verified the accuracy of MERRA-2’s results by comparing them to air quality measurements collected from the surface at U.S. embassies and consulates around the world. They analyzed PM2.5 exposure in 2019 to ensure that any changes in mortality associated with the COVID-19 pandemic did not influence the results.
There are multiple ways researchers can represent the size and shape of dust particles in MERRA-2 and other atmospheric models, and the research team found that estimates of PM2.5 deaths are more accurate if calculations are based on the aerodynamic size of dust particles rather than the geometric size.
“Aerodynamic size incorporates important information about the shape and density of dust particles that is relevant to how readily the particles fall out of the atmosphere and move into the respiratory system,” explained Yu. Though the geometric size for dust—which is larger than the aerodynamic size—is commonly used by atmospheric scientists, doing so in this type of health outcome research would lead to an overestimation in the number of deaths attributable to dust by about 1 million people, according to Yu.
Key sources of satellite data that were used to constrain MERRA-2 include the MODIS and Multi-angle Imaging SpectroRadiometer (MISR) sensors on NASA’s Terra and Aqua satellites. MERRA-2 covers the modern satellite era (1979 to present) and runs using the Goddard Earth Observing System (GEOS) model. In GEOS, airborne particles are simulated using the Goddard Chemistry Aerosol Radiation and Transport (GOCART) model.