Data stories · 9 min read
US air quality in October 2026 was a little worse than last year, after a late-month spell in the Midwest and East
US air quality, October 2026 vs 2025. Fine particles rose in 213 of 300 metro areas, led by Chattanooga, Lancaster, Rome and Toledo. Wenatchee improved most.
By the Good Air team
In this article · 6 sections
US air quality in October 2026 was slightly worse than in October 2025. Across 300 metro areas, the median level of fine particle pollution (PM2.5) for the whole month, 1 to 31 October, was 6.3 micrograms per cubic metre (µg/m³), up from 5.3 in October 2025, and 213 metro areas got dirtier while 87 got cleaner. Both years were well inside the Good range of the US Air Quality Index (AQI).
The gap opened late. Until 18 October the two years were almost identical. From 19 October particle levels rose across Indiana, Ohio, Michigan and Illinois, and to a lesser degree in the Southeast and the Northeast, and stayed high to the end of the month. Compared with October 2024, both years were cleaner.
These numbers are our own analysis of the EPA's AirNow monitor data, comparing only monitors that reported in both years. AirNow data are preliminary and have not yet been validated by the EPA.
Key findings
- A small rise. The median metro area went from 5.3 µg/m³ to 6.3, a US AQI equivalent of 29 to 34. Of 300 metro areas, 130 rose by 1 µg/m³ or more and 34 fell by 1 or more.
- Late October made the difference. The daily national median averaged 4.8 µg/m³ in 2025 and 4.9 in 2026 from 1 to 18 October, then 4.6 against 6.1 from 19 to 31 October.
- Biggest rises: the Midwest, the Southeast and California's Central Valley. Chattanooga (+5.3), Lancaster, Pennsylvania (+5.0), Rome, Georgia (+4.8) and Toledo (+4.8) rose most. Indianapolis rose 4.3 and Fresno 4.2.
- Biggest fall: Wenatchee, Washington (−6.7), where wildfire smoke filled the valley in October 2025. Arizona and Texas border towns, Phoenix, Los Angeles and Houston were also cleaner.
- Fewer very bad days, more Moderate ones. Across 906 monitors, days with a 24-hour average above 35.4 µg/m³, the line where the AQI passes 100, fell from 15 to 9. Monitor-days above 9.0, the top of Good, rose from 3,618 to 4,862, or from 13% to 18% of all readings.
- Cleaner than 2024. Among 778 monitors with data in all three years, the median was 6.9 µg/m³ in October 2024, 5.3 in 2025 and 6.0 in 2026.
Where did the air get worse, and where better?
The tables compare mean PM2.5 from 1 to 31 October in each year, at the same monitors. A metro area is the EPA's core-based statistical area as AirNow labels it, so "Fresno" means Fresno County. On the 2024 AQI scale for PM2.5, 0 to 9.0 µg/m³ is Good, 9.1 to 35.4 Moderate and 35.5 to 55.4 Unhealthy for Sensitive Groups (EPA, February 2024).
Biggest rises
| Rank | Metro area | Oct 2025 (µg/m³) | Oct 2026 (µg/m³) | Change | Monitors |
|---|---|---|---|---|---|
| 1 | Chattanooga, Tennessee | 5.4 | 10.6 | +5.3 | 1 |
| 2 | Lancaster, Pennsylvania | 5.2 | 10.2 | +5.0 | 2 |
| 3 | Rome, Georgia | 5.5 | 10.3 | +4.8 | 1 |
| 4 | Toledo, Ohio | 5.5 | 10.3 | +4.8 | 2 |
| 5 | Yuba City, California | 5.6 | 10.1 | +4.4 | 1 |
| 6 | Elkhart-Goshen, Indiana | 6.3 | 10.7 | +4.4 | 1 |
| 7 | Fort Wayne, Indiana | 6.1 | 10.4 | +4.3 | 1 |
| 8 | Indianapolis, Indiana | 6.5 | 10.7 | +4.3 | 2 |
| 9 | Fresno, California | 9.2 | 13.3 | +4.2 | 3 |
| 10 | Miami, Oklahoma | 9.2 | 13.2 | +3.9 | 1 |
Muncie, Indiana (+3.9) and Columbus, Ohio (+3.5) were close behind. Changes are ranked on unrounded values. Most metro areas in these tables have only one or two monitors that reported in both years, so one monitor can move the result. Chattanooga's single monitor missed 6 and 7 October 2026, so its 2026 mean covers 29 days.
Biggest falls
| Rank | Metro area | Oct 2025 (µg/m³) | Oct 2026 (µg/m³) | Change | Monitors |
|---|---|---|---|---|---|
| 1 | Wenatchee, Washington | 12.2 | 5.5 | −6.7 | 1 |
| 2 | Nogales, Arizona | 11.1 | 6.4 | −4.7 | 1 |
| 3 | Eagle Pass, Texas | 8.7 | 4.8 | −3.9 | 1 |
| 4 | Yuma, Arizona | 9.6 | 6.2 | −3.4 | 1 |
| 5 | Anchorage, Alaska | 7.6 | 4.7 | −2.9 | 1 |
| 6 | Moscow, Idaho | 7.4 | 4.7 | −2.7 | 1 |
| 7 | Des Moines, Iowa | 8.1 | 5.5 | −2.6 | 1 |
| 8 | Los Angeles, California | 8.9 | 6.6 | −2.3 | 9 |
Large metro areas
| Metro area | Oct 2025 | Oct 2026 | Change | Monitors |
|---|---|---|---|---|
| Indianapolis | 6.5 | 10.7 | +4.3 | 2 |
| Columbus, Ohio | 6.4 | 9.9 | +3.5 | 3 |
| Pittsburgh | 4.7 | 7.3 | +2.6 | 13 |
| Chicago | 6.3 | 8.9 | +2.6 | 20 |
| Detroit | 5.6 | 8.1 | +2.4 | 12 |
| Atlanta | 5.9 | 7.4 | +1.5 | 4 |
| San Francisco Bay Area | 4.9 | 6.3 | +1.4 | 6 |
| Sacramento | 5.8 | 7.1 | +1.3 | 8 |
| New York City | 4.4 | 5.7 | +1.3 | 17 |
| Dallas-Fort Worth | 8.1 | 8.2 | +0.1 | 4 |
| Boston | 4.2 | 4.2 | 0.0 | 11 |
| Austin | 7.7 | 6.3 | −1.4 | 4 |
| Houston | 9.4 | 7.7 | −1.7 | 5 |
| Phoenix | 7.1 | 5.1 | −2.1 | 9 |
| Los Angeles | 8.9 | 6.6 | −2.3 | 9 |
The highest October 2026 means were in the San Joaquin Valley: Hanford 15.3 µg/m³ and Fresno 13.3, both Moderate. For a ranking of cities by overall AQI this autumn, see the most polluted cities in the US.
Which regions changed?
Median of each region's monitors, 1 to 31 October:
| Region | Monitors | Oct 2025 (µg/m³) | Oct 2026 (µg/m³) |
|---|---|---|---|
| California Central Valley | 25 | 6.6 | 9.7 |
| Midwest and Ohio Valley (IL, IN, OH, MI, WI, KY) | 144 | 5.8 | 7.7 |
| Southeast | 124 | 5.2 | 6.3 |
| Northeast and Mid-Atlantic | 149 | 4.1 | 5.1 |
| Northern Rockies (MT, ID, WY) | 39 | 3.3 | 4.3 |
| Pacific Northwest (WA, OR) | 112 | 5.3 | 5.7 |
| Desert Southwest (AZ, NM, NV) | 44 | 4.8 | 4.2 |
| Texas | 37 | 7.4 | 6.7 |
| Southern California | 35 | 8.0 | 5.9 |
Our Central Valley group is the metro areas from Redding to Bakersfield; Southern California is Los Angeles, Riverside-San Bernardino, Oxnard-Ventura, San Diego and El Centro. The Southeast is Georgia, Alabama, Tennessee, the Carolinas, Florida, Mississippi, Virginia and West Virginia.
What changed, and why?
We only give a cause where an agency recorded one. Where none did, we say so.
The Midwest's late-October spell
From 1 to 18 October, the Midwest's daily median was lower in 2026 than in 2025 (5.7 µg/m³ against 6.3). From 19 to 31 October it averaged 10.2, against 4.4 a year earlier, reaching 12.5 on 23 October, 13.4 on 25 October and 12.9 on 28 October before falling back on 30 October. On 22 October 2025 it had been 2.7.
At single monitors it went higher. On 27 October two Indianapolis monitors averaged 39.4 and 37.0 µg/m³ over the day, in the Unhealthy for Sensitive Groups range, and a third averaged 36.7 on 28 October. The Indiana Department of Environmental Management (IDEM) declared a PM2.5 advisory for Marion County from 6 pm on 27 October to 10 am on 28 October, citing "observed and forecasted high concentration" of PM2.5 (NWS Indianapolis, 27 October). Toledo's Eastside monitor averaged 62.5 µg/m³ on 21 October, an AQI equivalent of 156 (Unhealthy), the highest October 2026 day at any US monitor in AirNow's daily files. Toledo's other monitor, Erie, read 30.5 that day, so both rose.
We can't say why. IDEM's corrected advisory gave no cause; a first version, issued 36 minutes earlier, had advised people to limit exposure to wildfire smoke (NWS Indianapolis, 27 October). NOAA's satellite analysts saw a broad area of light smoke over the Midwest on 22 October (NOAA Hazard Mapping System, 22 October), and on 27 October light smoke from fires in the Mississippi Valley stretching from Illinois to Michigan (NOAA, 27 October). NOAA's product does not say how much smoke reaches the ground, and we found no agency that linked the readings to it. We found no agency statement about the high readings of 28 and 29 October.
A year earlier the Midwest had its own short bump: its median reached 11.9 µg/m³ on 4 October 2025, when NOAA described very light smoke over the Midwest and Great Lakes, "mainly attributed" to farm burning along the Mississippi Valley and in the South-east (NOAA, 4 October 2025).
The Southeast and Northeast: a smaller late rise, cause unknown
The Southeast's median averaged 6.5 µg/m³ from 19 to 31 October 2026, against 4.6 a year earlier, and the Northeast's 5.7 against 3.8. Chattanooga's monitor read 16.1 to 16.8 from 13 to 16 October and 18.1 and 18.4 on 29 and 30 October. We found no agency statement that explains either rise.
California's Central Valley: smoke from Yosemite early in the month
The valley's median was 12.1 µg/m³ from 1 to 6 October 2026, against 4.8 a year earlier. The Fresno-Garland monitor averaged 17.5 to 19.0 from 1 to 4 October. In those days NOAA reported the Dome Fire in Yosemite National Park producing smoke "drifting northwestward towards the Central Valley" (NOAA, 1 October; 2 October). In mid-October it also reported fires in the northern Central Valley, where Yuba City and Chico are, producing local smoke (NOAA, 16 October).
After the smoke the gap narrowed: 8.4 µg/m³ from 7 to 31 October against 7.3, and from 19 to 31 October the valley was slightly cleaner than a year earlier (8.5 against 8.7). We can't say why. Yuba City's monitor was above 9.0 on 19 of 31 days in 2026, and on 5 in 2025. For why the San Joaquin Valley traps pollution, see why Fresno's air quality is so bad.
Wenatchee: last year's smoke, not this year's
Wenatchee improved most because October 2025 was smoky. Washington's Department of Ecology reported "unhealthy to hazardous" air in Wenatchee and Cashmere on 6 October 2025, with an air quality alert for Chelan and Douglas counties, and blamed the Labor Mountain Fire (Washington Smoke Blog, 6 October 2025; 10 October 2025). The metro area's compared monitor, at Leavenworth, averaged 35.1 to 53.6 µg/m³ from 6 to 8 October 2025. The Wenatchee-Fifth St monitor, which AirNow does not assign to a metro area, averaged 20.7 µg/m³ in October 2025 and 7.2 in 2026, with a peak of 79.2 on 6 October 2025.
Southern California, Arizona and Texas: cleaner, cause unknown
Los Angeles, Riverside-San Bernardino, Phoenix, Austin, Houston and the border towns of Nogales and Eagle Pass were all cleaner. We found no agency statement that explains this. Southern California had a strong Santa Ana wind event in mid-October 2026 (see our Santa Ana winds explainer), but we cannot tell from these data whether it helped. We did not analyse rainfall.
How our numbers compare with our own archive
Good Air also stores AirNow readings hour by hour. At six monitors (in Fresno, Denver, Indianapolis, Queens, Charlotte and Houston), AirNow's October 2026 means matched our archive within 0.2 µg/m³ on the days both have; Indianapolis Washington Park, for example, was 14.5 against 14.3. Single days differ more, typically by 0.5 to 1.9 µg/m³, because AirNow's days run midnight to midnight local standard time and ours run in UTC.
How we measured
- Data. AirNow's daily data files for 1 to 31 October 2024, 2025 and 2026, from the public archive at files.airnowtech.org, downloaded on 2 November 2026. We used the 24-hour average PM2.5 for each US monitor, which AirNow calculates midnight to midnight in local standard time (AirNow daily data fact sheet, August 2025). Government monitors only: no low-cost sensors or satellite estimates. The 31 October 2026 file was complete (1,145 monitors).
- Preliminary data. AirNow says its data "are not fully verified or validated" and should be treated as preliminary; final data come from the EPA's Air Quality System (AQS) later (same fact sheet). Numbers may shift a little when AQS data come out.
- Same monitors. A monitor counts only if it reported on at least 27 of the 31 days in both years. 906 permanent US monitors qualified, out of 1,155 reporting in October 2025 and 1,199 in October 2026. We left out 20 temporary monitors, such as wildfire response units, and one monitor with frozen readings (the same value five days running): Hamshire, near Beaumont, Texas, which reported 0.0 every day from 26 to 31 October 2025.
- Metro areas. Each monitor's mean over its valid days, then the mean of a metro area's monitors. 646 monitors are in one of 300 metro areas; the other 260 count only in the national and regional figures. 188 of the metro areas have a single monitor. Santa Rosa-Petaluma's only monitor, at Sebastopol, read zero or below on four days in October 2025, so its rise (2.0 to 5.8 µg/m³) may partly reflect the instrument.
- AQI equivalents convert a mean concentration with EPA's 2024 PM2.5 breakpoints, in force since 6 May 2024, when the top of Good fell from 12.0 to 9.0 µg/m³ (EPA, February 2024). They are not official daily AQI values.
- Causes come only from agency sources we opened: NOAA's Hazard Mapping System smoke text, Washington's smoke blog and an NWS-relayed IDEM alert. NOAA's smoke text describes smoke seen from space, not at ground level.
Live readings: all US cities. If smoke reaches you, our wildfire smoke safety guide covers what to do.
Frequently asked questions
Was US air quality worse in October 2026 than in October 2025?
Slightly. The median of 300 US metro areas was 6.3 µg/m³ of PM2.5 from 1 to 31 October 2026, against 5.3 in October 2025, and 213 metro areas got dirtier. Both are in the Good range. The difference came almost entirely after 18 October, when particle levels rose across the Midwest and, less sharply, the Southeast and Northeast.
Why was air quality bad in Indianapolis in late October 2026?
PM2.5 built up across Indiana, Ohio, Michigan and Illinois from 19 October. On 27 October two Indianapolis monitors averaged 39.4 and 37.0 µg/m³, and IDEM issued a PM2.5 advisory for Marion County through 10 am on 28 October. The final advisory gave no cause. NOAA saw light smoke over the region, but no agency has said how much of it reached the ground.
Was there wildfire smoke in the US in October 2026?
Yes, in places. NOAA reported smoke from the Dome Fire in Yosemite drifting towards California's Central Valley in early October, when the valley's PM2.5 was more than double last year's. Wenatchee, Washington, had much less smoke than in October 2025, when the Labor Mountain Fire caused unhealthy to hazardous air.
What is a good PM2.5 level?
On the US AQI, a 24-hour average of 0 to 9.0 µg/m³ is Good, 9.1 to 35.4 Moderate and 35.5 to 55.4 Unhealthy for Sensitive Groups. EPA lowered the top of Good from 12.0 to 9.0 in May 2024, so older reports can call the same air Good that is now Moderate.
Is AirNow data official?
No. AirNow publishes data in near real time, and says it is preliminary and not fully validated. The EPA's Air Quality System holds the final, quality-assured data, which come out months later. We will check these results against it when it is available.
