Why Your Air Purifier's Air Quality Reading Might Be Wrong, And How to Fix It

Most consumer air purifiers use optical sensors. A light source, either an LED or a laser, shines into a small chamber. Particles passing through scatter the light, and a detector measures this scattering to estimate PM2.5.

Two broad types show up in air purifiers: simpler infrared/LED sensors in many budget and mid-range models, and laser sensors (like the Plantower PMS series) in Xiaomi, Levoit, and several higher-end models. Infrared sensors generally have lower sensitivity below about 1 micron, while laser sensors detect down to about 0.3 microns. That doesn't mean laser sensors are more accurate in every situation. At least one independent lab comparison found accuracy on both types depends heavily on the particle being measured, with some laser sensors performing worse than infrared ones on certain dust. Both types can be affected by dust on the lens, inlet, or housing.

Dust on the lens or photodetector is one documented cause. As dust builds up, less light reaches the detector, and the sensor often causes under-reporting of pollution. Field studies on low-cost optical sensors have observed sensor response dropping over weeks as dust accumulates.

But dust isn't the only cause, and it doesn't always push readings the same way. Dust deposited on the sensor's optics over time tends to reduce sensitivity and cause under-reading, while heavy airborne dust during an active storm can produce erratic or overestimated readings. High humidity adds a separate effect: condensation directly on the sensor can make it show worse air than is actually present (Philips flags this explicitly in its manuals). In general, dry high-dust conditions push readings low; high humidity pushes them high. Accuracy also varies by particle composition and region independent of cleanliness: the same sensor type has been found overestimating PM2.5 in some Indian cities and underestimating it in others. Cleaning helps, but it's one factor among several.

Indian cities add their own version of this problem, not because the dust is unusually small, but because there's a lot of it. Source apportionment studies in Delhi have found wind-assisted resuspension of surface dust contributing over half of PM10 load in some years, and in Ghaziabad, in the Indo-Gangetic Plain, dust has been found to make up roughly 44 to 54% of PM10, rising further during the worst pollution days. That volume of dust likely clogs sensor openings and coats optical components faster than in lower-dust environments, regardless of the size of any individual particle.

Cleanliness aside, these sensors don't perform the same way across the full range of pollution levels.

At low concentrations, several common sensors are unreliable. One lab comparison found two widely used sensors correlated poorly with a reference monitor when PM2.5 was below 20 µg/m³. Erratic readings on a relatively clean day are often this known low-end limitation, not a dirty sensor. Worth noting: 20 µg/m³ is low by Indian standards but still above the WHO's 2021 annual guideline of 5 µg/m³ and its 24-hour guideline of 15 µg/m³.

At moderate to high concentrations, typical of an average Indo-Gangetic Plain day, sensors do best. The same study found six sensor models all responded well once PM2.5 rose above 50 µg/m³.

At very high concentrations, research doesn't agree on direction. A Beijing field study on a common Plantower sensor found it underestimated true PM2.5 by close to 25% above 250 µg/m³. At least one laboratory comparison of a related Plantower model observed a sharper non-linear response above around 400 µg/m³, with underestimation at high concentrations and the opposite bias at lower ones, though exact figures vary by model, aerosol type, and conditions. The Beijing study itself saw both effects on different days: underestimation during high-concentration episodes, overestimation once humidity rose above 60% (worse past 75%), and especially poor performance during dust storms.

The takeaway: many studies find a tendency toward underestimation on dry, high-concentration days, and overestimation on humid days or during dust storms. Indo-Gangetic Plain winters frequently involve both elevated dust loading and high humidity, making sensor behaviour less predictable than at other times of year. Treat a "very poor" or "severe" reading as a reliable signal to act on, and treat the precise figure with more skepticism the worse the air gets.

There's also a manufacturer-stated limit. The Plantower PMS7003, used in several consumer purifiers, lists an effective PM2.5 range of 0 to 500 µg/m³ (maximum error of ±10% from 100 to 500 µg/m³), with a separate, higher maximum range beyond that. Past 500, you're outside the range the manufacturer tested for.

Always turn off and unplug the purifier first. Locate the sensor cover (check your manual for its exact position, usually on the side or back of the unit). Open it gently, using a non-sharp tool if needed.

There's no single method that works for every brand, and whether a vacuum is safe on the sensor itself varies too: Coway permits it on the lowest power setting, while Philips' manuals explicitly say not to vacuum the particle sensor. Check your specific manual rather than assuming one method applies across brands.

Outside those specific exceptions, don't put water directly on a sensor, and don't use sharp objects to open covers.

It helps to clean the pre-filter at the same time, since a heavily loaded pre-filter can also drag down overall performance.

Even a freshly cleaned sensor won't match a reference-grade monitor exactly. Low-cost optical sensors have accuracy limits tied to humidity, particle size, and composition, not just cleanliness. If a reading still seems off after cleaning the sensor and pre-filter, that's more likely a sensor limitation than something further cleaning will fix.

For context on how different agencies report the same air differently, see India AQI vs US AQI. To cross-check your unit's reading against a nearby reference-grade CPCB station, use our AQI Checker.

If your air purifier's reading doesn't match outdoor conditions or how the air feels, clean the sensor and pre-filter before assuming the unit has failed. It's a five-minute check that often resolves the issue.

Pair sensor maintenance with:

• Proper purifier sizing (use our CADR Calculator)
• Regular pre-filter cleaning
• Ventilation when outdoor AQI allows it (see our Ventilation Guide)

As of June 2026. Cleaning steps are based on official manufacturer manuals for the specific models cited above. Always check your own unit's manual, since methods can vary by model even within the same brand.