Air Quality Apps: How to Track Pollution Alongside Pressure

An air quality app is useful for headache tracking only if you understand three things about the number it shows you: which pollutant is driving it, how far away the measurement was taken, and whether the figure is a real observation or a model estimate. The composite index most apps display collapses several very different pollutants into one score by reporting whichever is worst, which means an index of 90 from ozone on a hot still afternoon and an index of 90 from fine particulate during a smoke episode are different exposures that may affect you differently. Tracking the index alone will blur exactly the distinction you are trying to find.

Combining air quality with barometric pressure in one log is genuinely worth doing, because the two are related in ways that create both real effects and false ones. Stagnant high-pressure conditions trap pollutants; frontal passages clear them. So pollution and pressure move together often enough that either one can appear to explain your symptoms when the other is responsible.

This is a guide to reading these apps critically and building a log that can actually distinguish between the two.

What the index number hides

Most countries publish a composite air quality index. The details differ by jurisdiction, but the general approach is shared: measure several pollutants, convert each to a common scale, and report the highest one.

That last step is where information is lost. The headline number tells you the worst pollutant, not the overall air. An index of 75 could mean moderate ozone with everything else clean, or moderate fine particulate with everything else clean. Those are different atmospheric situations with different sources, different daily timing, and quite possibly different effects on you.

The pollutants usually included:

• PM2.5 — fine particles under 2.5 micrometres, from combustion, vehicles, wood burning and wildfires. Small enough to reach deep lung tissue and enter the bloodstream. Peaks in winter from heating in many regions, and during smoke episodes.
• PM10 — coarser particles including dust and pollen fragments. Peaks in dry windy conditions and during construction.
• Ozone — not emitted directly but formed photochemically from traffic and industrial emissions in sunlight. Peaks on hot, sunny, stagnant afternoons and is often worse downwind of a city than in it.
• Nitrogen dioxide — traffic-dominated, peaking at rush hours and falling off sharply within a couple of hundred metres of a busy road.
• Sulphur dioxide and carbon monoxide — industrial and combustion sources, generally lower in most modern urban areas than they once were.

Any app that lets you see the individual pollutant readings is worth more than one that shows only the composite. If the app you use will not break the index down, replace it.

Note also that indices are not comparable across countries. The same air can produce noticeably different index values under different national scales, because the breakpoints and averaging periods differ. If you travel, do not compare numbers between systems.

Where the number came from

This is the second thing most people never check, and it matters as much as the pollutant breakdown.

Reference monitors are regulatory-grade instruments operated by environmental agencies. They are accurate and well maintained, but sparse — a large city may have only a handful, and rural areas may have none for many miles. If the nearest one is thirty kilometres away and upwind, its reading may have little to do with the air at your front door.

Low-cost sensor networks are the dense crowdsourced grids that have transformed local air quality visibility over the past decade. There may be one on a street near you. They are excellent for showing spatial patterns and relative changes, but individual units drift, and most types read particulate high in humid conditions because water absorbed onto particles makes them appear larger. Good platforms apply a correction; not all apps that consume the data do.

Modelled estimates are what fills in everywhere with no sensor at all. The app interpolates from distant monitors, satellite data, emissions inventories and meteorology. This is legitimate science, but it is a calculation rather than a measurement, and its accuracy falls off sharply at the neighbourhood scale — precisely the scale at which traffic pollution varies most.

Many apps present all three identically, with no indication of which you are looking at. Find out how your app sources its data, and prefer one that names the station or sensor and shows its distance from you.

Timing, averaging and the trap in a daily figure

Pollutants have characteristic daily cycles, and an app showing a single daily figure smooths them away.

Nitrogen dioxide peaks with morning and evening traffic. Ozone peaks in mid to late afternoon after hours of sunlight, and falls to near zero overnight. Fine particulate in winter often peaks in the evening as domestic heating fires up and again in the early hours under a shallow nocturnal inversion.

If your headaches start in the late afternoon and your log records a daily average, you are recording a number that includes eight hours of clean overnight air. Log the reading at the time your symptoms began, not the day's summary. An app with an hourly history graph makes this straightforward; one showing only the current value does not.

Averaging periods also differ by pollutant within the same index — some are computed over an hour, some over eight, some over twenty-four — so a single index value may be describing different time windows depending on which pollutant is dominant that day.

How pressure and pollution interact

This is the part that makes a combined log worth keeping, and also the part that makes it easy to draw the wrong conclusion.

High pressure traps pollution. A stagnant anticyclone brings light winds, sinking air and often a temperature inversion — a layer of warm air sitting over cooler surface air that acts as a lid. Pollutants emitted at ground level accumulate beneath it instead of dispersing. Winter particulate episodes in valley cities and summer ozone episodes both occur overwhelmingly under high pressure.

Low pressure and fronts clear it. Wind, mixing and rain scavenge particles out of the air. Air quality typically improves sharply as a front passes.

The consequence is a genuine correlation between pressure and pollution in most places. And that creates a real problem for interpretation: if your headaches cluster during stagnant high-pressure spells, is it the pressure pattern, the accumulated particulate, the inversion's effect on humidity and temperature, or the fact that stagnant weather also tends to come with poor sleep in hot conditions?

You cannot answer that by tracking one variable. You can begin to answer it by tracking both and looking for the cases where they diverge:

• High pressure with clean air — a windy anticyclone, or a rural location with no local sources. If symptoms still occur, pollution is unlikely to be the driver.
• Low pressure with poor air — a wildfire smoke plume arriving ahead of a front, or a dust event on strong pre-frontal winds. If symptoms occur here, pollution is a stronger candidate.

Those divergent days are the informative ones. Most days tell you nothing because both variables move together. A log that runs long enough to accumulate a decent number of divergent days is worth far more than a short intensive one.

Building a log that can distinguish them

A workable structure, recorded at symptom onset rather than at the end of the day:

• Date and time of onset
• Severity and duration
• Barometric pressure and, more importantly, the change over the preceding six to twelve hours
• Individual pollutant readings, not just the index — at minimum PM2.5 and ozone
• Distance and direction to the sensor you used, noted once and kept consistent
• Whether you were indoors or outdoors, and for how long
• The standard confounders: sleep, hydration, caffeine, alcohol, stress, menstrual cycle, screen hours

The indoor-outdoor field matters more than people expect. Outdoor particulate penetrates buildings reasonably well; ozone is highly reactive and largely destroyed on indoor surfaces, so indoor ozone is typically a small fraction of outdoor. If you spent an ozone-alert afternoon inside with the windows closed, your exposure bears little relation to the outdoor number.

Pressure Pal handles the pressure half of this — current barometric pressure and the trend, logged automatically alongside your symptom entries — so the manual effort is limited to adding the pollutant figures. If you want a location-specific view of the pressure side, the city forecast pages show the local pattern and what typically drives it, which is useful context when a stagnant spell sets in.

Keep it for three months minimum before drawing conclusions. Headache frequency is noisy enough that shorter series mostly measure chance.

Choosing an app

What to look for:

• Individual pollutant breakdowns, not just a composite index
• Named source stations with distance, and an indication of whether the value is measured or modelled
• Hourly history, so you can retrieve the reading for the hour your symptoms started
• Forecast as well as current conditions, which is what makes the app actionable rather than merely descriptive
• Data export, if you intend to do any analysis
• Clear labelling of the index standard in use

What to be wary of:

• Apps that show a single number with no breakdown and no source
• Apps whose forecast extends many days out with implied precision that models do not support
• Any product that promises to eliminate the health effects of pollution rather than to report it

Government environmental agency apps tend to be conservative, well sourced and honest about uncertainty, though sometimes sparse. Crowdsourced platforms give far better spatial resolution at the cost of individual sensor reliability. Using one of each and noting when they disagree is more informative than trusting either alone.

What to do with a bad reading

Tracking is only worth the effort if it changes behaviour. Reasonable responses when levels are elevated:

• Particulate high: close windows, run a HEPA purifier in the room you occupy most, postpone outdoor exercise. Exercise matters because exertion multiplies the volume of air you breathe.
• Ozone high: stay indoors during the afternoon peak. Indoor ozone is far lower than outdoor, so this is unusually effective. Shift outdoor activity to morning.
• Nitrogen dioxide high: move away from busy roads. Concentrations fall steeply with distance, so a route one street back from a main road is meaningfully cleaner.
• All elevated during a smoke episode: windows closed, filtration on, and check the forecast for when the pattern breaks — usually with the next frontal passage, which your pressure log will also show.

Frequently asked questions

Which matters more for headaches, air quality or barometric pressure? There is no general answer, and that is the reason to track both. The published evidence for pressure as a migraine trigger is larger and longer-standing than that for pollution, but individual variation dominates in both cases. Your own log is the only source that answers the question for you.

My app shows a different number from the government site. Which is right? Probably neither is wrong; they are likely using different sources, different averaging periods or different index standards. Pick one and stay with it, because consistency matters far more than which you chose.

Does rain clean the air? It removes particulate effectively through wet deposition. Its effect on gases is smaller. Rain also usually comes with wind and mixing, both of which help, so air quality generally improves after rain.

Why is ozone higher in the suburbs than downtown? Ozone forms downwind as precursor emissions react in sunlight over hours, and nitric oxide from heavy traffic actually destroys ozone locally. The result is often lower ozone in the traffic-dense centre and higher levels downwind.

Should I track the index or individual pollutants? Individual pollutants. The index tells you only which one is worst, and it is the specific pollutant that carries the useful information.

Do indoor air quality monitors help with this? They answer a different and often more relevant question, since most exposure happens indoors. A combination of an outdoor data source and an indoor monitor gives a much fuller picture than either alone.

How long before a combined log tells me anything? Three months at minimum, and longer is better, because the informative days are those where pressure and pollution diverge, and those are a minority.

---

This article is general information, not medical advice. If you have persistent or worsening headaches, speak with a healthcare professional.