Air Quality Index Explained: What AQI Means for Your Head
AQI is not a measurement. It is a piecewise conversion of pollutant concentrations onto a common 0 to 500 scale, reporting only the single worst pollutant at that moment, using breakpoints that differ from country to country. Understanding those three facts changes how you read the number — and explains why an AQI of 80 can feel very different on two different days.
For anyone whose headaches respond to air quality, the index is genuinely useful. It is also routinely over-interpreted. People treat it as a linear measure of how dirty the air is, compare readings between countries as though the scales matched, and assume the number describes the air they are currently breathing rather than an average over a window that has already closed.
None of that makes AQI a bad tool. It makes it a tool worth learning to read properly, particularly alongside a pressure forecast, because the two variables often move in opposite directions.
What the index actually does
Air quality monitors measure concentrations of individual pollutants in physical units — micrograms per cubic metre for particulate matter, parts per billion for gases. Those numbers are not comparable to each other. A concentration of 35 for PM2.5 and a concentration of 35 for ozone mean entirely different things about health risk.
The index solves that by mapping each pollutant onto a common scale using health-based breakpoints. Each pollutant has its own set of concentration ranges, and each range maps linearly onto a corresponding index range. Within a segment the relationship is linear; between segments the slope changes. That is what "piecewise" means, and it has a consequence people rarely notice.
The same index step does not correspond to the same concentration step at different parts of the scale. Moving from an index value of 50 to 100 involves a much smaller concentration increase than moving from 150 to 200. The scale is deliberately stretched at the clean end so that small changes near the threshold of concern are visible, and compressed at the dirty end where everything is bad anyway. Treating AQI as a linear ruler will mislead you about how much the air actually changed.
The pollutants typically included are fine particulate matter under 2.5 microns, coarse particulate under 10 microns, ground-level ozone, nitrogen dioxide, sulphur dioxide and carbon monoxide.
The dominant pollutant rule
This is the single most important thing to understand about AQI, and the least well known.
The reported AQI is the highest of the individual pollutant sub-indices, not an average of them. If PM2.5 gives 45, ozone gives 92 and nitrogen dioxide gives 30, the reported AQI is 92 and the dominant pollutant is ozone. The other two contribute nothing to the headline figure.
Several consequences follow.
The number tells you nothing about what you are actually breathing unless you also look at which pollutant is driving it. Most reporting services show this, usually in smaller text, and it is worth more attention than the number itself.
Different pollutants act through different mechanisms. Fine particulate crosses into the bloodstream and drives systemic inflammatory responses. Ozone is a powerful airway irritant that acts on mucosal surfaces. Nitrogen dioxide is a traffic marker and an irritant. If your headaches track particulate, an ozone-driven AQI of 110 may not affect you much, and vice versa. You may have a personal sensitivity to one pollutant and not others, and the headline number cannot show you that.
The dominant pollutant follows the weather in predictable ways. Ozone is photochemical: it requires sunlight and precursor emissions, so it peaks on hot, sunny, stagnant afternoons in summer, typically mid-afternoon to early evening. Particulate accumulates whenever vertical mixing is suppressed, which means winter inversions, still nights, and any period of persistent high pressure. Wintertime AQI is usually particulate-driven and summertime AQI is often ozone-driven, in the same city.
Two days at AQI 90 can therefore be entirely different exposures, and if you are logging your symptoms it is worth recording the dominant pollutant alongside the number.
The category bands
The US scale, which is the one most commonly encountered in English-language reporting, uses six bands.
0 to 50, Good. Air quality poses little or no risk. Most sensitive people find nothing here.
51 to 100, Moderate. Acceptable, with a caveat for unusually sensitive individuals. This band is wider than people assume, and the upper end of it is where many headache-sensitive people start noticing something. If your log shows a threshold, it is quite often somewhere in the 70s or 80s rather than at the official 100.
101 to 150, Unhealthy for Sensitive Groups. This is the band named for you. It includes people with respiratory or cardiac conditions, older adults, children — and reasonably, anyone with a documented symptom response. The standard advice is to reduce prolonged or heavy outdoor exertion.
151 to 200, Unhealthy. Everyone may begin to experience effects; sensitive groups more seriously. Outdoor exertion should be limited and moved indoors where possible.
201 to 300, Very Unhealthy. Health alert territory. Sensitive people should remain indoors with filtration.
301 and above, Hazardous. Emergency conditions, typically seen during major wildfire smoke events.
The band boundaries are population-level advisories, not personal thresholds. They are set from epidemiological evidence about groups. Your own response may begin well below the sensitive-group line, and a personal threshold identified from your own logs is more useful to you than the official one. This is worth stating clearly, because people frequently dismiss their own consistent observation on the grounds that the number was "only moderate".
AQI is not the same number everywhere
Comparing an AQI reading in one country to a reading in another is close to meaningless without knowing which scale is in use.
The United States uses the 0 to 500 index described above, with breakpoints set by the national air quality standards.
Europe reports an index built on different bands with different reference concentrations, and it is generally more conservative for particulate — the same air can produce a worse-sounding category in Europe than in the US.
India, China, Canada and Australia all use their own schemes, with different pollutant sets, different breakpoints and in Canada's case a different structure entirely, based on a combined health-risk formulation rather than a worst-pollutant rule.
Consumer apps and low-cost sensor networks may apply the US formula to data from any country, or convert between scales, or report raw concentrations relabelled as an index. Two apps open on the same phone can disagree substantially about the same location for entirely mundane reasons.
The practical rule: pick one source and one scale, and stay with it. Your own threshold is only meaningful within a consistent reporting system. Switching sources mid-log will introduce a step change that looks like a real signal and is not.
Timing, lag and spatial resolution
AQI is more retrospective than it appears.
It is an average over a window. Particulate sub-indices are conventionally based on 24-hour averages, and ozone on 8-hour averages, though many services also publish shorter-interval or "NowCast" style estimates that weight recent hours more heavily during rapidly changing conditions. Either way, a headline number is describing a period rather than an instant.
There is a reporting delay. Monitoring data goes through validation and aggregation before publication, so what you see may be an hour or more behind the atmosphere. During a fast-moving event — a smoke plume arriving, a frontal passage clearing the air — the published figure lags reality noticeably.
Monitors are sparse. Regulatory stations are expensive and there may be only a handful across a metropolitan area. Your reading is interpolated from the nearest one, which could be many kilometres away, in a different microclimate, on the other side of a ridge, upwind or downwind of a motorway. Local variation in particulate near roads is substantial.
Forecast AQI and observed AQI are different products. Forecasts model tomorrow's conditions and carry real uncertainty, particularly for ozone, which is sensitive to how hot and sunny the day actually turns out.
For symptom logging, the sensible response is to record the observed value at a consistent time of day rather than checking repeatedly and recording the worst number you saw.
Using AQI alongside a pressure forecast
This is where the two variables become genuinely complementary, because the meteorology that produces bad air is largely the opposite of the meteorology that produces pressure triggers.
Persistent high pressure means poor air and a flat barometer. Subsidence within a ridge warms the air aloft and creates an inversion that caps the boundary layer. Vertical mixing stops, winds go light, and pollutants emitted at the surface accumulate underneath, day after day. The barometer is high and steady — the pattern most weather-sensitive people read as safe — while AQI climbs. Winter valley smog and summer ozone episodes both live here. The pressure trend for an inland basin location shows these long flat ridges clearly, and so does a desert location in summer.
Frontal passage means moving pressure and clearing air. The approaching system brings wind and often rain, which disperses and scavenges accumulated pollutants. AQI improves sharply, sometimes within hours, while the barometer falls and then rises. For a pressure-sensitive person, this is the risky configuration and the air is fine.
The overlap cases are the interesting ones. Dry gusty pre-frontal winds can raise coarse particulate from soil and roads even as pressure is falling. Wildfire smoke can be transported into a region under any synoptic pattern and decouples the usual relationship entirely. And a stagnant ridge that finally breaks down gives you the pressure change and the pollutant flush at the same time.
So when the two indicators disagree, that is information rather than noise. A bad day with clean air and a moving barometer points to pressure. A bad day with a flat barometer and elevated AQI points to air quality. Over a month of logging both, the pattern separates cleanly for most people, and the answer determines what you actually do — filtration and indoor time, or acute migraine planning and threshold protection. Checking a pressure forecast for your location alongside a single AQI reading takes about twenty seconds a day and is the whole method.
Frequently asked questions
What AQI level should trigger action if I get headaches?
Official guidance puts the sensitive-group threshold at 101, but many people who track their own symptoms find a personal threshold in the 70s or 80s. Use the official band as a starting assumption and let your log refine it. Exertion multiplies your inhaled dose, so the threshold for outdoor exercise should be lower than for ordinary activity.
Why do two apps show different AQI for the same place?
Different source monitors, different interpolation, different averaging windows, different national scales, or low-cost sensor data rather than regulatory data. Pick one source and stay with it so your own readings remain comparable over time.
Does a good AQI mean the air is safe for me?
Not necessarily. The index covers a specific set of regulated outdoor pollutants. Pollen, mould spores, indoor allergens and volatile organic compounds are not in it at all, and indoor air can be considerably worse than outdoor air while AQI reads good.
Which pollutant is most associated with headache?
The evidence for short-term increases in headache and migraine presentations is most consistent for fine particulate, PM2.5, which is small enough to enter the bloodstream and drive systemic inflammation. Ozone and nitrogen dioxide have supporting evidence as airway and mucosal irritants. Individual sensitivity varies, which is why recording the dominant pollutant is worth the effort.
Can I rely on AQI to predict a bad day?
Partly. Forecast AQI carries real uncertainty, particularly for ozone. It is best used as one input among several rather than as a standalone prediction, and it works considerably better in combination with a pressure trend than alone.
Why do I feel unwell when AQI is fine but the weather is changing?
Because they are different mechanisms. Barometric change acts through pressure differentials and, plausibly, central sensitisation; air quality acts through inflammation and irritation. Clean air during a frontal passage is exactly the situation where a pressure trigger shows up unobscured.
Does an air purifier make AQI irrelevant indoors?
It helps substantially for particulate, which is what most high-AQI days are made of, provided the unit is HEPA-rated and sized for the room. It does far less for ozone and gaseous pollutants unless it includes activated carbon, and it cannot help while windows are open.
Pressure Pal tracks barometric pressure and its trend for your location, which is the second half of the picture when an air quality reading alone does not explain how the day went.