One app says AQI 82. Another says 117. A nearby sensor says 145. Which number is right?
Sometimes one source is stale or poorly configured. More often, the screens are answering different questions with different pollutants, locations, instruments, averaging periods, or AQI systems. Comparing only the final number hides those differences.
Use this checklist before deciding that an app is wrong.
1. Are the apps using the same AQI system?
“AQI” is not one worldwide formula. Governments and providers can use different pollutant sets, concentration breakpoints, averaging periods, and category labels.
Google's official explanation of how air-quality indexes are calculated highlights these differences across countries. A numerical value on a European, Indian, Chinese, or proprietary index should not be compared directly with a U.S. AQI unless the provider explicitly identifies the selected scale.
Check the app's settings and legend for labels such as:
- U.S. AQI or EPA AQI;
- European Air Quality Index;
- India National AQI;
- a provider-specific index; or
- raw PM2.5 in µg/m³ rather than an AQI.
If the scales differ, switch them to the same index before comparing—or compare the underlying pollutant concentration instead.
2. Are they reporting the same pollutant?
The U.S. overall AQI is the highest of the pollutant-specific sub-indexes. EPA establishes AQI calculations for ozone, particle pollution, carbon monoxide, nitrogen dioxide, and sulfur dioxide. A screen that reports only PM2.5 can differ from a source whose overall AQI is driven by ozone or PM10.
This is common in regions where ozone is important but consumer sensors measure particles. South Coast AQMD lists missing pollutants as one reason private apps can disagree with its local AQI map. See the agency's AQI methodology and FAQ.
Open the detail view and look for dominant pollutant, main pollutant, or individual sub-indexes. Do not assume that every colored dot represents overall AQI.
3. Do the readings describe the same place?
A government monitor may be several miles away. A PurpleAir sensor may sit on the next block. A weather app may choose the nearest station, a city-center point, or a model grid cell.
Pollution can vary over short distances near roads, grills, chimneys, construction, valleys, coastlines, and smoke boundaries. Google's station-selection documentation warns that the nearest station does not necessarily represent the user's exact location and that not every station measures every pollutant.
Check:
- the coordinates or station name;
- distance from your location;
- whether the source is indoors or outdoors;
- nearby local emission sources; and
- terrain and wind during smoke events.
Hyperlocal does not automatically mean more representative. A sensor beside a barbecue can accurately measure that tiny plume while poorly representing the neighborhood.
4. Are the timestamps and averaging periods the same?
AirNow's current particle and ozone AQI uses the NowCast, which relates recent hourly observations to the AQI scale and responds when conditions change. AirNow also notes that most data appear after the measured hour rather than instantaneously. See AirNow's data timing and validation notes.
PurpleAir's map can display different averaging periods, including a default ten-minute view documented in the official PurpleAir Map guide. A fast sensor average can jump when smoke arrives while a NowCast or hourly model changes more gradually.
Compare:
- measurement time, not only the phone's current time;
- ten-minute, one-hour, NowCast, and 24-hour values;
- current observation versus today's forecast; and
- whether one app has not refreshed.
Our current AQI and forecast guide explains why those labels are not interchangeable.
5. Is one value measured and another modeled?
An app may show:
- a regulatory or reference monitor;
- a low-cost optical sensor;
- a model estimate between monitors;
- a blend of monitors, sensors, satellites, weather, traffic, and smoke data; or
- a forecast.
Google describes its model as a fusion of governmental monitors, sensor networks, atmospheric models, satellite information, traffic, land cover, and meteorology. AirNow primarily displays observations submitted by monitoring agencies for public AQI reporting. PurpleAir's own map displays its sensor network.
All can be useful, but they have different strengths. A reference monitor offers standardized measurement quality at one location. A dense sensor network can reveal local PM2.5 variation. A model can fill gaps but may miss a very local event.
6. Has low-cost sensor data been corrected?
Optical particle sensors infer particle mass from light scattering. Humidity, aerosol composition, particle density, aging, and placement can affect the result.
PurpleAir lets users select data layers, averaging periods, and conversions. Its calibration documentation explains that different conversions address different conditions. Changing that setting can change the displayed AQI.
The AirNow Fire and Smoke Map includes PurpleAir sensor data after quality screening and a U.S.-wide correction. AirNow's Fire and Smoke Map questions and answers explains that corrected sensors appear alongside permanent and temporary monitors.
When comparing PurpleAir with AirNow, do not assume that an uncorrected PurpleAir map layer and the corrected Fire and Smoke Map should match exactly.
7. Are both numbers observations?
A forecast is a prediction of future air quality. A NowCast-based current AQI uses recent observations. A final daily AQI uses the complete required daily data. Apps sometimes place these next to one another under short labels such as “today.”
Open the label before comparing. If one number is a forecast, disagreement with the current reading is not a data conflict; it means observed conditions differ from the predicted daily pattern so far.
A two-minute diagnostic workflow
When two apps disagree:
- Set both to the same AQI system.
- Confirm the same pollutant, ideally PM2.5 concentration in µg/m³ for smoke.
- Match the location or compare nearby stations on a map.
- Match timestamps and averaging periods.
- Identify monitor, sensor, model, or forecast.
- Check sensor correction and quality indicators.
- Look at several nearby points and the recent trend.
- Use official health guidance for the decision you need to make.
In the United States, EPA recommends AirNow for AQI and health information while recognizing that a local sensor can add short-term information at a specific location. The EPA sensor guidance explains the complementary roles.
Which source should you use?
Choose by task:
- Official U.S. current and forecast context: start with AirNow.
- Wildfire monitors, corrected sensors, fires, and smoke layers: use the AirNow Fire and Smoke Map.
- Hyperlocal PM2.5 sensor trends: inspect multiple PurpleAir sensors with appropriate settings.
- International or model-based coverage: review the provider's selected index, pollutant, source, and timestamp.
Our Air Quality App Finder compares documented use cases without claiming that one provider is universally best.
Before using any reading in a cigarette-equivalent calculator
Do not enter an unknown overall AQI simply because it came from a familiar app. First verify that it is:
- on the U.S. AQI scale; and
- specifically a PM2.5 AQI.
If the app gives raw PM2.5 in µg/m³, use the PM2.5 Cigarette Calculator. If it gives an ozone-driven or unknown overall AQI, do not convert it. Follow the full AQI conversion eligibility checklist.
Bottom line
Different AQI numbers often reflect different scales, pollutants, places, time windows, instruments, corrections, or models. The responsible comparison is not number versus number; it is method versus method.
Match those inputs first. Then use the source that fits your decision and follow the health guidance attached to the relevant pollutant and AQI category.