PSI bands explained: what the index covers and misses

The PSI is not a measurement. It is an index assembled from six separate pollutant measurements, each converted onto a common scale running from 0 to 500. NEA calculates a sub-index for every one of the six, then publishes the highest of them as the PSI.

That last step is the one most people skip past. The published figure is not a blend of six pollutants. It is the worst of six, wearing the name of the index. Five sub-indices can sit low while the sixth on its own sets the number that appears on the app.

The six are fine particles (PM2.5), coarse particles (PM10), sulphur dioxide, nitrogen dioxide, ozone and carbon monoxide. PM2.5 was the last to join, folded into the computation in 2014. Each carries its own averaging period, and those periods are not the same length.

Ozone carries a second rule on top. Once the 8-hour concentration passes 785 micrograms per cubic metre, NEA switches that sub-index to the 1-hour concentration instead. The exception matters less than the shape it reveals. This is a set of separate rules stitched onto one scale, not a single instrument reading.

• •PM2.5, PM10 and sulphur dioxide are each averaged over 24 hours
• •Ozone and carbon monoxide are each averaged over 8 hours
• •Nitrogen dioxide is averaged over 1 hour, and reported only at 1,130 µg/m³ or higher

NEA names five bands and fixes the boundaries at whole numbers. Anything above 300 falls in the top band, and the computation scale itself stops at 500.

The top band is wide by design. The four bands below it span fifty or a hundred index points each, while Hazardous absorbs everything from 301 upwards. Two readings can share that label and describe very different air. Past 300 the band name has stopped separating anything, and the concentration underneath is the only part still carrying information.

The right-hand column below is the part worth keeping. Because the index reports whichever sub-index runs highest, a haze figure of 120 nearly always means the 24-hour fine-particle concentration sat between 56 and 150 micrograms per cubic metre. The band name travels through news coverage and weather apps. The concentration underneath it rarely does.

The boundaries are also where the reporting changes, not where the air does. A move from 100 to 101 flips the label from Moderate to Unhealthy while the fine-particle concentration behind it shifts by roughly one microgram per cubic metre. The label jumps and the air barely moves. Bands exist to make a scale communicable, so a figure sitting close to a boundary is better read as close to a boundary than as a member of a category.

NEA attaches health guidance to each band, and a separate personal guide is keyed to the 1-hour fine-particle concentration. Those publications are where to read what a band means for a person. This page stops at how the figure is assembled and what it can support.

The index covers a rolling 24-hour period. Every hourly update recalculates the average across the day just gone, so a figure read at three in the afternoon still carries last night inside it. Nothing about that is a defect. Describing a whole day is the job the index was given.

NEA settled on that period deliberately. The health research the boundaries are calibrated against studied concentrations averaged over 24 hours rather than hourly ones. An index that chased hourly spikes would no longer line up with the evidence its own thresholds came from.

The result is a lag, and it runs in both directions. A sharp deterioration after lunch enters an average still weighted by twenty-three calmer hours, so the figure creeps up while the sky changes quickly. For most of the first day of a spell, the index sits below what a person standing outside would guess.

Rough numbers show how large the effect gets. Smoke arriving at noon and holding steady has fed only a handful of hours into the window by the evening news. Everything else in that window is the clear air that preceded it, pinning the figure down. The index reaches the level the afternoon deserved roughly a day later, by which point the wind may already have turned.

This is why a figure quoted in the morning is a poor guide to an afternoon plan. It is an accurate account of the day behind it and a weak prediction of the hours ahead. NEA covers that separately during haze, publishing a forecast range for the next 24 hours rather than leaving the backward-looking figure to do a job it cannot.

Since the index publishes the maximum sub-index, the headline belongs to a single pollutant at any given hour. Reading it as a general score for the air loses that. Two afternoons both sitting at 90 can be two unrelated situations.

Through a transboundary haze spell, the fine-particle sub-index is the one setting the figure. Smoke carried in from regional burning loads that fraction far harder than it loads the other five. That is why NEA publishes the 1-hour fine-particle concentration during those stretches, and why haze coverage quotes it.

Outside a spell the picture is less predictable. Ozone builds through a hot, still afternoon under strong sun. Carbon monoxide and nitrogen dioxide follow traffic and combustion. A moderate figure in clear weather is usually a different pollutant from a moderate figure in smoke, and the index alone will not tell you which.

Every figure covered so far describes outdoor air. The monitors sit outside, and NEA reports them across five regions: north, south, east, west and central. A regional figure is an average over a large area, so the air at one block can differ from the number pinned to the region it falls in.

The distance between outdoor and indoor is wider again. Air inside a shut flat is not the outdoor reading with a delay applied. It is a mix of what has leaked in, what has settled onto surfaces, and what is being produced inside by cooking and cleaning. None of that reaches an outdoor monitor.

Closing a flat up changes the indoor air on its own terms, and not only in the direction people intend. Sealing slows the arrival of outdoor particles. It also holds indoor sources in and lets carbon dioxide climb in an occupied room. Indoor CO2 has its own page here.

How tightly a home is closed decides how much of the outdoor figure ever arrives in the room. Every gap in the building fabric is a route for outside air, and every extractor fan left running is a pump drawing more of it through. Two flats on one corridor, sharing a regional figure, can hold quite different air.

So the reading works as context rather than as an instruction. It reports what the outdoor air has been doing across a region, and roughly how long the current spell has run. What is in your bedroom sits outside its reach, and always has.