AIRI, Atmospheric Interdisciplinary Research Institute
Air Quality Investigation · Summer 2026

Same Air,
Different Story

We checked the top 10 AQI websites during real air quality events this summer. Each told a very different story about how safe the air was to breathe.

TWO EVENTS · TEN SITES · HOURLY SCREENSHOTS

We picked ten AQI sites by searching Google for "AQI now" and taking the top ten results: AirNow, AirNow's Fire and Smoke Map, Google's own AQI panel, IQAir, AQICN, PurpleAir, AccuWeather, MSN Weather, Weatherbug, and Wunderground. These are not necessarily a list of "the best" tools. It's just what anyone searching for information on their air quality might have found.

We tracked those same ten sites, hour by hour, through two very different real-world events this summer: the Boyle Heights Warehouse Fire in Los Angeles in June, and the "World's Largest Fireworks Event" on July 4th at Washington DC's National Mall. Same method, two very different kinds of events. Pick a tab below to see what each one showed.

PURPLEAIR · SENSOR NETWORK
PurpleAir map of Los Angeles on June 21, 2026 at 9:05am showing a dense cluster of dark red and maroon dots over the city center, indicating Very Unhealthy air quality readings up to 202 AQI.
AIRNOW · REGULATORY MONITORS
AirNow map of the same LA region at the same moment, showing the city center as solid green with no alert, and its only orange/red zone displaced far to the east near Ontario.

SAME MOMENT · SUNDAY, JUNE 21, 2026 · 9:05 AM PDT

10 sites we tracked. Not a curated "best of" list, just the top ten Google results for "AQI now"
0.4% to 13% of hours each site rated the air unhealthy for sensitive groups or worse, over the same three days, depending which site you checked
202 peak AQI PurpleAir's dense sensors showed downtown at 9am Sunday. Other sites read that same hour very differently

If you'd checked the air quality in LA that Sunday morning, what you saw depended a lot on which site you visited. PurpleAir showed a plume of Very Unhealthy smoke sitting right over the city center. AirNow's own map, at that exact same moment, showed that same patch of city as a calm, uniform green, with the only bad air over 30 miles away near Ontario.

Neither one is wrong. They're just built to do different jobs.

Watching the plume move

PurpleAir's readings come from thousands of privately and publicly owned sensors packed tightly across the city. That's dense enough that stepping through the hours in order doesn't just show a number changing. It shows smoke actually drifting across the basin overnight, then thinning out by afternoon. AirNow's official monitors are much more spread out and built to accurately determine multi-pollutant concentrations as well as track broad regional trends, but it's harder to see a moving plume with a sparser network, so the same event doesn't register as well on this map.

Animated comparison of PurpleAir and AirNow maps over 24 hours from the evening of June 20 through June 21, 2026, showing a smoke plume visibly building overnight and receding by afternoon on PurpleAir's dense sensor map, while AirNow's sparser official map shows comparatively little change.

Same 24 hours, both sources · Saturday 6pm through Sunday 6pm PDT

Hourly AQI category mix, June 19 to 21, 2026, Pacific time
Each bar is what percent of that site's map area was colored in each AQI category that hour, not a single point reading. Hover to compare an exact hour across all ten sites.
View the underlying hourly numbers as a table

Why the same air gets different stories

Regulatory monitors, like the ones behind AirNow, are the most accurate single readings you can get. They're diligently and carefully maintained by state and local agencies, and provide the data that triggers official air quality alerts. The tradeoff compared to PurpleAir is coverage. There are typically only a handful of regulatory monitors across any given metro area because of their high maintenance and complexity, so a plume that's intense on one block, whether from an urban fire or a fireworks show, can be less visible on the regional map if the main plume never reaches one of these monitors.

PurpleAir provides a different approach. Thousands of low-cost sensors that can be installed in and around people's homes. In some areas, the number of sensors is dense enough to trace a plume block by block or catch one right where it starts. But individually, sensors like these can read high in heavy particulate smoke and at high humidity since both scatter light and can cause the optical sensor to read that as more particulate. Combining these dense sensor networks and high-quality regulatory information can provide an advantage. This is one of the reasons why there is a PurpleAir correction factor, so the data between AirNow and PurpleAir can be combined and compared.

Combining dense sensors and regulatory monitors is an advantage of the AirNow's Fire and Smoke Map. It's a separate site from the standard AirNow map (one of the ten we tracked), and it blends both signals: regulatory monitors plus a corrected version of PurpleAir and Clarity sensor data. Wide coverage and local detail, together.

In both events we tracked, it caught more of the spike, sooner, than the standalone AirNow map did. Applying the correction factor allows both sets of measurements to be show together. If you only add one extra site to your AQI routine during a smoke event or a pollution event, the AirNow Fire & Smoke Map is the one to use.

Every website presents AQI in a different way. Beyond the raw data, sites differ in how they color the map, how often they refresh, and how they interpolate between stations. So even two sites reading similar underlying data can end up showing a visibly different picture. During the LA warehouse fire event, AccuWeather barely moved out of "Moderate" the entire three days, even while other sites swung from clear to hazardous and back. During the DC fireworks, the opposite occurred where AQI from PurpleAir and Accuweather were elevated well before the show started.

Additionally, we don't just see that the numbers differ, we find that the AQI vocabulary can be different. AccuWeather calls its mildest bad-air tier "Poor," in orange. Weatherbug also uses the word "Poor," but for its most severe tier, the range AirNow calls "Very Unhealthy" and "Hazardous." The same word on two different websites can be describing air that's mildly worse than average or potentially dangerous. Here's how all ten sites map their own categories onto the same six EPA severity tiers:

A table comparing how each of the ten AQI sites labels and colors the same six EPA severity tiers (I through VI). AirNow, AQICN, and Google use the standard Good/Moderate/USG/Unhealthy/Very Unhealthy/Hazardous scale, while AccuWeather uses Excellent/Fair/Poor/Unhealthy/Very Unhealthy/Dangerous, Weatherbug uses Excellent/Fair/Moderate/Low/Poor with Poor spanning the two worst tiers, and Wunderground labels the third tier 'Unhealthy if Sensitive' in a distinct peach color.

What to actually do about it

  • Check more than one source, especially during a fire, a big fireworks show, or any other smoke event. If you only add one, make it AirNow's Fire and Smoke Map. It's built to combine regulatory accuracy with dense sensor coverage.
  • The worst air doesn't always line up with when you'd expect. Smoke can build overnight and clear by afternoon; firework smoke can peak an hour or two after the show ends. Don't trust that a single check will hold for the rest of the day.
  • If you're pregnant, have kids, elderly, are living with asthma, COPD, or heart disease, be conscious of when air quality is Moderate, Unhealthy for Sensitive Groups or higher. When any source flags smoke nearby, wildfire, urban, or otherwise, it's worth limiting outdoor time even if another source still looks clear.