TOI Explains: AQI of 3,000? Why your app data seems more alarming than the official one

Gurgaon: On Diwali night, social media posts showing app readings of 3,000 AQI (air quality index) spread so widely that, on Monday morning, a Delhi resident got a call from a friend in the US. "What are you still doing in Delhi? Your AQI is 3,000," the friend asked.Readings of 1,000, 2,000 and even 3,000 are often shared during severe pollution episodes, which read very different from Central Pollution Control Board's (CPCB) official data. This does not make the official number inaccurate, nor the app readings necessarily false. The figures cannot be compared because the official data is an index and app readings are actual value at that point of time at the station it is taking its data from.Concentration vs indexConcentration is the directly measured value of a pollutant, in micrograms per cubic metre. AQI is an index calculated from that concentration of different pollutants using prescribed breakpoints, and has no unit."AQI is calculated separately for each pollutant at each location, using its measured concentration and the prescribed breakpoint values. The highest pollutant-wise sub-index determines the overall AQI, and that pollutant is called the dominant pollutant. Concentration is the directly measured value, expressed in micrograms per cubic metre. AQI is an index calculated from that concentration and has no unit," Dr Mohan P George, former additional director at Delhi Pollution Control Committee (DPCC), told TOI.

How CPCB calculates AQICPCB's national AQI considers eight pollutants: PM2.5, PM10, nitrogen dioxide, sulphur dioxide, carbon monoxide, ozone, ammonia and lead. Each concentration is converted into a sub-index using health-based breakpoint ranges. The highest sub-index sets the overall AQI.An overall AQI needs data for at least three pollutants, including PM2.5 or PM10, and at least 16 hours of data for each pollutant sub-index. If data is insufficient, available sub-indices may still be published, but the overall AQI may not be calculated.Particulate matter sub-indices use 24-hour average concentrations. Carbon monoxide and ozone use eight-hour averages. CPCB's real-time system captures continuous station data and uses running averages. A 6am AQI, for instance, incorporates data from 6am the previous day to 6am that day.CPCB's NCR portal also shows hourly PM2.5 and PM10 concentration trends, which are not the same as the overall AQI. It notes that real-time data arrive from stations without prior validation, which is done later. CPCB's national AQI dashboard gives station-wise and city-level readings, and its bulletin cautions that a city average may not represent variations across individual locations.India's official AQI scale ends at 500, but pollutant concentrations and health risks do not stop rising at that point. Some private platforms use other indices, such as the US AQI, or shorter averaging periods, which can produce values above 500."Unusually high AQI readings of 1,000 to 3,000 should not automatically be dismissed as incorrect, but they should not be directly compared with CPCB's AQI.," said Dr Vipul Chandani, senior associate in air quality sector at Centre for Study of Science, Technology and Policy. "Such readings may reflect a different AQI methodology, averaging period, or extrapolation beyond the highest category of an index. Platforms such as IQAir report US AQI⁺ values in India, which use hourly average concentrations, while India's AQI uses 24-hour moving averages for particulate matter. Different concentration breakpoints and averaging periods can therefore produce substantially different numbers. The actual pollutant concentration, measurement source and calculation method are essential to understanding these figures.""India's AQI scale being capped at 500 does not mean actual pollution concentrations or associated health risks cannot increase further. During extreme pollution episodes, citizens should look beyond the AQI and examine actual PM2.5 and PM10 concentrations as well. Reporting both the AQI and pollutant concentrations provides a more complete understanding of pollution severity and potential exposure," he added.Equipment and methods also matterThe number of pollutants measured, the calibration of instruments and the method used to convert concentrations into an index all differ between sources. So does the equipment."In reference-grade equipment, air is drawn through a filter, and the particles are deposited on the filter, tape or other media. Their mass is then determined through beta-ray attenuation. The equipment uses standardised conditions for air collection and measurement," Dr George said."Low-cost sensors generally count the particles passing through a tube and convert that count into a particulate matter concentration using a conversion factor. However, particles are not always the same size, and humidity can change their size. Water vapour may also be counted as particles, which affects the conversion. It is not always clear where different apps source their data or which formula they use. The numbers may differ, but what matters is understanding the pollution category and the reliability of the underlying data. I am not of the opinion that one source's data is wrong and another's is correct. It depends on how the data is generated and the methodology used," he added.What does a reading of 3,000 mean?"Apps report different readings because their methods differ. CPCB AQI is based on multiple parameters; some platforms calculate using selected pollutants. There are seven or eight parameters; pollutant with highest health impact is considered for AQI, could be ozone or PM2.5. Other platforms are taking some of the numbers and reporting the AQI. That may not be correct, but ours is the most correct one," said Dr DS Saha, former head of CPCB's air cell."People should not panic just because an app shows a figure like 3,000 AQI. It is important to understand the scale being used, how the AQI has been calculated, and which averaging period has been considered. If the scale has range of 0 to 50, 0 to 500 or 0 to 5000, it is also necessary to understand," Dr George added.When an app reports an AQI far above CPCB's reading, users should check whether it uses India's AQI or another scale, which pollutant is driving the figure, whether the reading is hourly or a longer average, the actual PM2.5 or PM10 concentration, the monitoring location and timestamp and whether the data come from reference-grade equipment or a low-cost sensor.

A reading above 500 should neither be dismissed automatically nor compared directly with CPCB's index.