Why do two air-quality sources show different numbers?

Before deciding that an air-quality app is wrong, identify what its number represents. An AQI score, a pollutant concentration, yesterday’s daily result and a forecast can all appear under an “air quality” heading. Their values are not interchangeable.

The useful first question is whether the two sources describe the same measurement. If they do not, subtracting their numbers will not measure the disagreement.

Three panels, three different claims

Fictional exercise: imagine that three panels refer to the same invented town. The values below are teaching material, not a current air-quality report.

“A reports 50 more pollution units than B” would be an unsupported rewrite of this table. AQI points are not a mass concentration, and these two panels cover different periods. C’s unit describes a concentration; its forecast and grid location add two more differences.

To investigate A, ask for another source’s PM2.5 result for station S and the same observation time, with the same AQI method. If only a nearby station is available, describe that as a spatial comparison. Do not quietly relabel it as an instrument check at the same point.

Find the scale before interpreting the colour

The U.S. AQI assigns values and colour categories to air-quality conditions. A concentration such as PM2.5 in micrograms per cubic metre describes a different quantity. Converting between the two requires the appropriate pollutant, averaging period and current calculation rules.

Do not identify an index from the country name on a screenshot. An app displaying a city abroad may use the U.S. scale or another system. Look for the index’s name and methodology link in the source itself.

The European Air Quality Index currently uses hourly pollutant concentrations and offers separate station and modelled layers. It also marks some gap-filled station values. Its categories should not be read as if they were numbers on an unnamed universal scale. Save the selected layer and the legend with the result.

“Current” still needs a time definition

AirNow’s guide to daily AQI, forecasts and NowCast distinguishes them explicitly. Daily particle AQI concerns a 24-hour period. The particle NowCast uses recent hours to describe the latest conditions, with the averaging responding to how quickly conditions change. A forecast is a prediction for its stated period.

Keep the observation time, its time zone and the page retrieval time as separate fields. Two screenshots taken at 10:00 can display observations from different hours. The word “updated” may refer to the page rather than the sample; open the details before assigning it a meaning.

Do not average pollutant scores into a new index

Suppose a second invented record has already calculated valid U.S. pollutant AQI values of 70 for PM2.5 and 95 for ozone, and those are the pollutants included in the exercise. The reported overall value is the larger one, 95. Taking their average, 82.5, would create a number the reporting rule did not request.

The EPA’s May 2026 technical assistance document explains the highest-pollutant rule and the calculation from concentrations. “Highest” here concerns the calculated index values, not whichever pollutant happens to have the largest raw number in its own units. Missing a pollutant does not establish that its concentration was zero.

Leave a useful discrepancy for the next reviewer

Capture the original URL and screenshot, location or station identifier, pollutant, concentration and unit if supplied, index name, averaging method, observation period and zone, retrieval time, and whether the value is observed, modelled, gap-filled or forecast. Record absent details as unknown.

If two properly matched records still differ, preserve both. Check source revisions, sensor corrections and provider explanations before attributing the difference to an error or deception. AirNow says its public data do not undergo the full certification used for its regulatory database; see About the Data. A saved value may later be revised.

For our three panels, the finding is “these values are not a matched comparison,” followed by the exact missing observation needed. Use the OSINT Jet investigation brief to make that request specific, and keep the source details in the report template. For decisions about exposure, follow the current guidance from the responsible local authority; this exercise checks the meaning of evidence.