Filling the air pollution data gap

The combined effects of outdoor and indoor air pollution contribute 6.7 million Every year, premature deaths make air pollution one of the world’s worst Leading risk factors For mortality and chronic diseases, according to the World Health Organization and the Institute for Global Health Metrics and Evaluation. Yet for most people, esp Low and middle income countriesAccess local air quality information Mostly limited.

„I often see 'world’s most polluted city’ rankings, but these kinds of rankings can be misleading,” said Randall Martin, an atmospheric scientist at Washington University in St. Louis. „The density of air quality monitors around the world is so low that no one usually knows which city is the most polluted, and even less so on a given day. Despite the recent development of underground monitors, millions of people still live in countries that do not collect or publicly share high-quality air pollution data.

A team of NASA-funded researchers is trying to address the data shortage by developing a different way to monitor air pollution — powered by satellite observations, modeling and supercomputing. As part of an effort to make this approach more effective for cities in developing countries, the NASA team World Resources Institute (WRI) in a pilot project CityAQ.

Most countries have Limited or no access Expensive but accurate „note” Air quality monitoring stations to measure Fine particle material (Evening_2.5) and other harmful pollutants etc Nitrogen dioxide (No₂), Ozone (Oh₃), and Sulfur dioxide (So₂) lack of monitoring has consequences. A 2022 analysis In air pollution data shared by US embassies in 40 countries, cities given access to real-time air quality data saw significant drops in air pollution levels.

Even in large cities, highly reliable air quality monitors are scarce. By An assessment5,500 PM are available_2.5 Monitors 1,700 cities worldwide with populations of 300,000 or more, with an average of three monitors per city. In sub-Saharan Africa, there is approximately one monitor for every 16 million people; In India, everyone has a monitor 7 million people.

To address the deficit, Christoph Keller and Emma Noland and colleagues at NASA Office of Global Modeling and AssimilationMorgan State University and Harvard University have developed a global model regression and forecasting the composition of the atmosphere, including many common air pollutants. instrument, the Goddard Earth Observing System Composition Forecast (Geos-cf), including real-time weather observations and insights from multi-year satellite observations and several historical ranges Emissions Inventory. It produces five-day forecasts for the entire atmosphere each day.

Example animations from GEOS-CF—showing the global distribution of terrestrial PM_2.5 And oh₃ For September and October 2023 areas – shown above and below. During that period, massive wildfires raged in Canada, large dust storms swept through the Sahara Desert and parts of South Asia, industrial pollution accumulated in the Indo-Gangetic Plain and East Asia, and seasonal fires burned in South America and sub-Saharan Africa.

Geos-cf It is a version of a comprehensive modeling and data integration system Goddard Earth Observing System (GEOS) has been used by NASA scientists for decades for weather and climate research. The model calculates the concentration of air pollutants in 25-square-kilometer grid boxes at 72 levels in the atmosphere from the surface to the stratosphere—as well as several atmospheric components. As part of the forecasts, GEOS injects Millions Daily weather observations. It works with help Discover the supercomputer at NASA Climate Simulation Center at the Goddard Space Flight Center.

„The power of GEOS-CF is that it can be used to begin filling in the gaps between ground monitors, making GEOS-CF a viable information source for countries that do not have much access to ground-based air quality monitoring,” Noland said.

In GEOS-CF Constraints and Forecasts, Bogotá, Colombia, encompasses four of the 25-square kilometer grid boxes. that is enough Spatial resolution For GEOS-CF to represent and predict large-scale, widespread air pollution, such as smoke blankets that pollute urban air during crop fires in January and February.

The image below was captured Virus (Visible Infrared Imaging Radiometer Suite) at NASA-NOAA Finland NPP The satellite shows an example of a seasonal fire burning south and east of the city on January 28, 2022, before winds shifted and pushed the smoke toward Bogotá.

However, the spatial resolution of the data produced by GEOS-CF is generally insufficient to provide useful information at the neighborhood, block or street level, Keller explained. In some cases, cities need data from a few kilometers or less to properly assess the health risks of locating a school near a highway. activities Restricting traffic or promoting use etc Electric vehicles Or bicycles are effective ways to reduce pollution.

The NASA team and WRI launched the CityAQ project in 2020 to develop more localized and tailored air quality monitoring information for specific cities and neighborhoods. It works by ingesting sparse ground and air observation data. Machine learning A technique to improve the accuracy of GEOS-CF model output. As part of the project, WRI partnered with eight cities to share data from ground observers with NASA: Addis Ababa, Ethiopia; Jakarta, Indonesia; Kigali, Rwanda; Leon-Salamanca-Zelaya Metro, Mexico; Monterrey Metro, Mexico; Guadalajara Metro, Mexico; Sao Paulo, Brazil; and Bogotá, Colombia.

„Once we started training the model with ground data, we started getting even better forecasts,” Keller said. Got sample output from NASA’s Global Modeling and Assimilation Office website and WRIs Resource Watch Browser. Some cities etc Monterey and Bogotá also began hosting CityAQ data on city websites.

The chart at the top of the page shows an example of CityAQ forecasts for nitrogen dioxide and ozone, released on November 4, 2023, after the model was trained using information from Bogotá’s Carvajal neighborhood. Given the facility’s location near a major highway and industrial sites, it typically records the highest annual NO.₂ The concentration of any observatory in the city is and continues to be high World Health Organization standards NO to both₂ and PM_2.5. The second chart shows ozone (O₃) and PM_2.5 Release to the nearby Kennedy neighborhood.

In both cases, the model accurately captured concentrations of pollutants and captured subtle nighttime peaks that demonstrate the accuracy of model output, Keller said. The model output is accurate enough to add Bogotá CityAQ data to the mix of tools it uses to create air quality alerts for the city’s 7 million residents.

However, Bogotá had more robust ground data than many of the CityAQ pilot cities, which made it easier for the NASA team to improve the GEOS-CF predictions. Researchers are currently trying to find new ways to integrate the observations Low cost air quality sensorsSometimes the only option in developing countries—in GEOS-CF forecasts. Low-cost sensors are easy to operate and cost from a few hundred to a few thousand US dollars each, while reference monitoring stations cost more than $10,000 each and require expertise in calibration and data interpretation.

As part of a new initiative that began in 2021, the World Resources InstituteThe Clean Air Fundand other partners are developing a system called CanAIRy that will fine-tune outputs from reference monitoring stations and low-cost sensor networks with air quality monitoring data from GEOS-CF.

“CanAIRy Alert includes air quality forecasts developed for cities as part of the CityAQ project and includes cities such as Kigali, Rwanda; Nairobi, Kenya; Kampala, Uganda; and Accra, Ghana,” said Beatrice Cardenas, WRI’s global director for air quality. “But we are in the process of recruiting and building more cities Communities of practice To share information about how to access and use the tools. Our goal Register Many more cities.”

1. Explore the data




2. NASA Global Modeling and Assimilation Office (2023) GMAO CityAQ Locations.




3. NASA Global Modeling and Assimilation Office (2023) Geosmix Predictions Chart.




4. Resource Watch (2023) CityAQ datasets.




5. World Resources Institute (2023) CanAIRy alert.

NASA Earth Observatory image by Michala Garrison using data from NASA Office of Global Modeling and Assimilation and the World Resources Institute CityAQ Try. videos Joseph Artisone and Katie Jepson, courtesy of NASA Science Visualization Studio. Supercomputer photo courtesy of NASA Center for Climate Simulation. Story by Adam Voyland.