See how air quality changed day by day in a U.S. metro area.
Try the project ↓Help
Outdoor monitors measure several common pollutants. The EPA turns those readings into a daily Air Quality Index so people can see when measured air quality raised more concern.
Open its highest July day, then choose any other date in the daily calendar.
Select a month to see how many reported days were Good, Moderate, or above an index of 100.
See which measured pollutant produced the highest index on each day in the selected month.
Compare its middle day and rough-day mark with 438 other well-covered areas.
See how air quality changed day by day in a U.S. metro area.
Outdoor monitors turn several pollutants into one daily Air Quality Index. The higher the number, the greater the concern in the EPA scale.
The colored and patterned tiles follow the EPA categories. Choose a date to see the index, the leading pollutant, and how many monitoring sites reported.
Saturday, July 5
The daily index is the highest available pollutant index reported across sites in this area. It does not describe every block or every hour.
Pick a tile. The number is the day of the month.
How to read this: A higher tile value means greater concern on the EPA scale. Color and pattern show the category; the day card names the pollutant that produced the highest index.
Each row counts reported days by category. Choose a month to update its pollutant view and the daily calendar above.
EPA calculates an index for each measured pollutant. The highest one becomes the area’s daily index. These bars count which pollutant led in July.
Each dot is one of 439 areas with at least 300 reported days. Right means a higher middle day. Up means a higher rough-day mark.
The rough-day mark is the index that nine out of ten reported days were at or below. The highest tenth were above it.
How to read this: The middle day describes a typical reported day. The rough-day mark focuses on the higher end without letting one extreme day control the comparison. Differences can reflect pollution, weather, geography, and what monitors measured; this project does not separate those causes.
The 2025 file is complete under EPA’s reporting schedule. Every view uses a checked-in snapshot, so the experience makes no request to EPA after the page loads.
Why this exists
How did daily air quality change through 2025 in a chosen U.S. metro or micropolitan area, which pollutant most often set the daily index, and how did the area's typical and rougher days compare with areas that had similar monitoring coverage?
Choose one of 439 areas with at least 300 reported days, inspect any day in the 2025 calendar, select a month, compare the pollutant that set each daily index, and place the area's middle and rougher days beside the national distribution.
Method and limits
I selected the U.S. Environmental Protection Agency's pre-generated Air Quality System files through the allowlisted Data.gov catalog. Data.gov identifies the publisher as the EPA Office of Air and Radiation and links the EPA Standard Open Data License. That license says EPA-produced data is public domain by default. I opened the underlying EPA AirData download page and its report definitions before using the file.
I retrieved `daily_aqi_by_cbsa_2025.zip` on October 7, 2026. EPA marked the file as updated June 25, 2026. The 893,830-byte archive has SHA-256 `0e26b7ee20c0792700d1a0cf47ba0a892ec24a7e78596f138692026cf290058d` and contains one 13,068,348-byte CSV with 170,306 daily records for 498 core-based statistical areas. I used 2025 instead of the incomplete 2026 file because EPA says current-year Air Quality Index summaries remain preliminary until May 1 of the following year.
The reproducible Python reducer verifies the archive hash, byte counts, member name, columns, row count, area count, category labels, and pollutant labels. It retains the 439 areas with at least 300 reported days, leaving 158,758 daily records. That coverage rule keeps very sparse years out of the national comparison. It does not make monitoring identical across areas. The reducer corrects two corrupted display glyphs in the source names for Cañon City and Española; their area codes and measurements do not change. Each retained day stores its index, EPA category, main pollutant, and number of reporting sites in a compact checked-in string. The reducer repeatedly produced the same 1,299,775-byte snapshot with SHA-256 `7760a0cb6681e3a65fffc9d88cfc89709f6f2f677d79627211f0873cdac9267e`. The page imports this snapshot and makes no runtime request.
The middle reported day is the median daily index. The rough-day mark uses the nearest-rank 90th percentile: the value at position ceiling(0.9 × reported days) after sorting. It describes the higher end of the year without letting one extreme day determine the comparison. Month counts use EPA's published categories without combining Good and Moderate days. The daily main pollutant is whichever measured pollutant produced the highest available index across reporting sites, following EPA's definition.
These are area-level monitoring records, not direct readings for every neighborhood, hour, or person. Areas monitor different pollutant mixes at different numbers of sites, and a core-based statistical area may cover several counties or states. The views do not measure individual exposure, explain why an index rose, establish health outcomes, or rank places by overall livability. Weather, wildfire smoke, emissions, terrain, and monitoring coverage may all affect a pattern; this project does not separate those causes.