Air Quality from Space: How Satellites Map NO₂, PM2.5, and Other Pollutants

When COVID-19 lockdowns emptied streets and shuttered factories in early 2020, TROPOMI captured what might be the most vivid satellite images of 2020 — not photographs, but maps of nitrogen dioxide over China, Europe, and the United States. Cities that normally appeared as bright red hotspots on NO₂ maps turned blue within weeks. It was pollution made visible from space, responding in near-real-time to human activity.

Those images did more to demonstrate the capability of satellite air quality monitoring than years of scientific publications. They showed that satellites don't just measure air quality — they reveal its direct connection to what we do on the ground.

What Satellites Can Measure

Nitrogen Dioxide (NO₂)

The best-measured pollutant from space. NO₂ has strong absorption features in the blue-violet region (400-465 nm), and its atmospheric lifetime is short (hours to a day), meaning elevated concentrations stay close to their sources.

TROPOMI retrieves NO₂ by comparing the measured spectrum against a reference clean-atmosphere spectrum. The difference reveals the total column abundance of NO₂ — the total amount in the atmospheric column from surface to satellite.

What the maps show: Bright spots over cities (vehicle emissions), shipping lanes (marine fuel), power plants, and industrial complexes. Background levels are low; sources are clearly identifiable.

Sulfur Dioxide (SO₂)

Detected through UV absorption features. Major sources visible from space include:

• Coal-fired power plants (reducing globally due to emission controls)
• Metal smelters
• Volcanic eruptions (massive plumes trackable across thousands of kilometers)
• Shipping emissions

Carbon Monoxide (CO)

Measured by TROPOMI in the SWIR region. CO has a longer atmospheric lifetime (weeks), so it spreads further from sources. Satellite CO maps reveal:

• Biomass burning emissions (wildfires, agricultural burning)
• Industrial emissions
• Long-range transport of pollution

Formaldehyde (HCHO)

A proxy for volatile organic compound (VOC) emissions. Detected in the UV-visible region. Biogenic emissions from forests (isoprene) and anthropogenic sources (petrochemical industry) are distinguishable by their spatial patterns.

Aerosol Optical Depth (AOD)

Not a gas measurement but a column-integrated measure of aerosol loading (particles in the atmosphere). MODIS and VIIRS retrieve AOD from the visible reflectance of the atmosphere:

• High AOD: Hazy, polluted conditions (dust storms, biomass burning, industrial smog)
• Low AOD: Clean air

AOD is the satellite measurement most closely related to surface PM2.5, though the relationship depends on particle properties, vertical distribution, and meteorological conditions.

From Column to Surface: The Translation Problem

Satellites measure total atmospheric columns — the integrated abundance of a pollutant through the entire atmosphere. What people breathe is the surface-level concentration. Converting one to the other requires:

Chemical transport models: Simulate the vertical distribution of pollutants based on meteorology, emissions inventories, and atmospheric chemistry. The model provides the fraction of the total column that's near the surface.

Statistical relationships: Correlate satellite column measurements with co-located ground monitor data to build empirical conversion functions. This approach works well where ground monitors are dense (North America, Europe) but poorly where they're sparse (Africa, much of Asia).

Hybrid approaches: Combine model-predicted vertical profiles with satellite column observations to produce surface concentration estimates. These typically achieve R² values of 0.5-0.7 against ground monitors for daily estimates, improving to 0.7-0.9 for monthly averages.

Applications

Urban Air Quality Monitoring

Satellite NO₂ maps reveal intracity pollution patterns:

• Highway corridors produce elevated NO₂ lines
• Industrial zones appear as bright spots
• Downwind dispersion follows prevailing wind patterns
• Weekend-weekday differences are detectable (reduced traffic)

This spatial detail complements ground monitors, which measure accurately at a point but can't capture the spatial variability across a city.

Health Exposure Assessment

The World Health Organization estimates that 4.2 million premature deaths annually are attributable to ambient air pollution, predominantly PM2.5. Estimating health impacts requires population-level exposure assessment:

Satellite-derived PM2.5 estimates provide global coverage at ~1 km resolution (the van Donkelaar dataset, widely used in health studies). While less accurate than ground monitors at any single point, they capture the spatial variation that health studies need — especially in developing countries where ground monitoring is sparse or nonexistent.

Emission Inventory Verification

Countries report emissions to international bodies under climate and pollution agreements. Satellite measurements provide independent verification:

• China's SO₂ reduction: Satellite SO₂ data confirmed that China's coal plant scrubber installations reduced SO₂ emissions by ~75% between 2007 and 2017
• India's NO₂ trends: Increasing industrialization and vehicle fleet growth visible as rising NO₂
• Ship emission controls: The 2020 IMO sulfur regulations produced a measurable decrease in SO₂ along major shipping lanes

Wildfire Smoke Tracking

Biomass burning produces CO, NO₂, aerosols, and formaldehyde. Satellite monitoring tracks:

• Smoke plume extent and trajectory
• Downwind air quality impacts on populated areas
• Total emissions estimation for fire management

Limitations

Spatial resolution: TROPOMI's ~5 km pixels are too coarse for street-level air quality assessment. They can't resolve the difference between a busy intersection and a nearby park.

Temporal coverage: One overpass per day (around 13:30 local time for TROPOMI). Air quality varies throughout the day — morning rush hour, afternoon photochemistry, evening inversions. One snapshot doesn't capture this diurnal cycle.

Cloud interference: Like all passive optical sensors, TROPOMI can't see through clouds. In persistently cloudy regions, satellite air quality data has significant gaps.

Surface-level conversion: The column-to-surface conversion introduces substantial uncertainty, particularly for PM2.5 where the relationship between AOD and surface concentration depends on aerosol vertical profile, hygroscopic growth, and mixing height — all of which vary in time and space.

No direct PM2.5: Satellites measure optical effects of particles (AOD), not mass concentration directly. The conversion depends on assumptions about particle size, composition, and vertical distribution.

The Future

Geostationary air quality satellites: GEMS (Korea, launched 2020), TEMPO (US, launched 2023), and Sentinel-4 (Europe, planned) provide hourly observations over their respective continents — capturing diurnal air quality cycles for the first time from space.

Higher resolution: Next-generation instruments aim for 1-2 km resolution, approaching the scale needed for intracity pollution management.

Multi-pollutant synergies: Combining NO₂, SO₂, HCHO, CO, and AOD from a single platform enables comprehensive air quality characterization and source attribution.

Satellite air quality monitoring has evolved from a scientific research tool to a practical complement to ground-based monitoring networks. It can't replace ground monitors for regulatory compliance (the accuracy isn't sufficient), but it provides something ground monitors never will: wall-to-wall spatial coverage showing where pollution comes from, where it goes, and how it's changing.