Delhi AQI Today: Why Apps Show Different Numbers

Delhi AQI apps can disagree without either being fake. Learn how India AQI uses stations, rolling averages, pollutant sub-indices and data-quality rules.

By RawColumn Editorial ·

Delhi AQI Today: Why Apps Show Different Numbers

“Delhi AQI,” “live AQI” and “Delhi AQI today” are part of a 500,000-plus India search cluster in the current Google Trends window. Open two apps, however, and you may see different numbers for what appears to be the same place and time. That does not automatically mean one app is lying. The readings may come from different stations, update times, pollutants, averaging windows or AQI systems.

This explainer shows how India’s National Air Quality Index is constructed, why a city average can conceal neighbourhood variation, and what to check before repeating a “live” number. It does not diagnose an individual or replace a local public-health advisory. People with symptoms or existing heart and lung conditions should follow their clinician’s advice and official alerts.

An AQI is an index, not a direct pollutant measurement

A monitor measures concentrations: for particulate matter, usually micrograms per cubic metre; for gases, units appropriate to the pollutant and method. AQI converts those concentrations into a common, dimensionless scale with categories intended for public communication. The number 250 is therefore not “250 units of pollution.” It is an index value produced by a defined breakpoint calculation.

India’s National AQI uses six public categories: Good (0–50), Satisfactory (51–100), Moderate (101–200), Poor (201–300), Very Poor (301–400) and Severe (401–500). CPCB attaches broad possible health-impact language to each band. The category helps communication, but the underlying pollutant and exposure duration still matter.

The worst pollutant sub-index sets the station AQI

CPCB’s methodology calculates a separate sub-index for each available pollutant by locating its concentration between published breakpoints. The overall AQI at that monitoring location is the highest—or worst—of those pollutant sub-indices. This is why a station can have one “prominent pollutant” even while several pollutants are measured.

The rule avoids averaging a severe particulate signal away with low values for unrelated gases. It also means the same overall AQI can arise from different pollutant mixtures. An AQI of 220 dominated by PM2.5 is not scientifically identical to 220 dominated by ozone, even if both fall in the Poor category.

“Live” usually contains many hours of history

CPCB states that pollutant sub-indices generally use 24-hour average concentrations, with eight-hour averages for carbon monoxide and ozone. Its web system presents running averages: a displayed value at 6 a.m., for example, incorporates observations reaching back into the previous day. “Live AQI” is therefore a continuously updated rolling summary, not a one-minute snapshot of the air entering your lungs.

This has an important consequence. A short pollution spike may take time to influence the index, and a cleaner wind shift may take time to pull a high running average down. Some private services show near-real-time sensor concentrations or forecasts beside an AQI. If the averaging window differs, the numbers should not be expected to match.

Enough valid data must exist

Under CPCB’s published method, an overall AQI is calculated only when at least three pollutants have data, and one must be PM2.5 or PM10. A minimum amount of data—described as 16 hours for a sub-index—is also needed. When those conditions fail, the responsible display is “insufficient data,” not an invented city number.

Missing data can come from maintenance, calibration, communications failures or quality-control review. A blank station does not prove clean air, and an inactive monitor should not be replaced with the previous value without a clear timestamp. Always check whether the app exposes station status and last update.

Why two Delhi locations can disagree

Air pollution varies across space. Road traffic, construction, industrial activity, waste burning, cooking and heating, dust, vegetation and local topography can affect emissions. Wind speed and direction, temperature inversions, humidity and boundary-layer height affect dispersion and particle behaviour. A roadside station and a residential background station can legitimately report different conditions.

CPCB’s Delhi-NCR page explicitly notes that it uses an average across multiple monitoring locations for a simple regional indication while acknowledging that averaging is not scientifically sound because air quality varies by location. For personal interpretation, the nearest station is useful—but proximity alone is not enough. Consider whether the station setting resembles where you actually are.

Why apps can disagree even when they use official data

  • Different station: one app chooses the nearest monitor; another shows a city or NCR average.
  • Different timestamp: caches and update schedules can lag by minutes or hours.
  • Different AQI system: India’s AQI breakpoints are not the same as the US AQI or every private scale.
  • Different pollutant set: one service may lack a pollutant that becomes the worst sub-index elsewhere.
  • Different averaging window: a rolling 24-hour regulatory value can differ from a short-term sensor reading.
  • Forecast versus observation: modelled future air quality is not a monitor measurement.
  • Correction and validation: services can apply quality checks or retain provisional data differently.

Before comparing, open each service’s details and align location, station, time zone, update time, AQI standard, pollutant and whether the value is observed or forecast. If those fields do not match, the headline numbers are answering different questions.

PM2.5 and PM10 are particle size classes

PM10 refers to inhalable particles with aerodynamic diameter up to about 10 micrometres; PM2.5 is the finer fraction up to about 2.5 micrometres. They are mixtures, not single chemicals. WHO describes particulate matter as including components such as sulphates, nitrates, ammonia, sodium chloride, black carbon, mineral dust and water.

Coarse particles can include dust, pollen and sea spray, while fine particles often arise directly from combustion or form in the atmosphere from precursor gases. The categories overlap: PM2.5 is contained within PM10, so the two numbers should not be added together. Their measurement and health evidence differ, and “dusty” visibility alone cannot tell you the PM2.5 concentration.

AQI categories and WHO guidelines are not the same thing

India’s AQI bands are a national communication system based on pollutant breakpoints. WHO’s Global Air Quality Guidelines provide health-based concentration recommendations over defined averaging periods. WHO’s 2021 values include 5 µg/m³ annual and 15 µg/m³ 24-hour guidance for PM2.5, and 15 µg/m³ annual and 45 µg/m³ 24-hour guidance for PM10.

Do not place a WHO concentration beside an AQI number and compare them numerically. One is a pollutant concentration with an averaging period; the other is an index derived through breakpoints. WHO guidance is not an app-specific replacement AQI and is not itself India’s legal ambient-air standard.

How the breakpoint calculation works

For a pollutant concentration that falls between two concentration breakpoints, the method maps it linearly between the corresponding AQI breakpoints. In simplified form:

Pollutant sub-index = lower AQI breakpoint + (concentration − lower concentration breakpoint) × (AQI range ÷ concentration range)

The exact pollutant table, averaging method, units and rounding rules matter. Do not calculate from a screenshot unless you know which standard and concentration units it uses. The educational value of the formula is to show that AQI is transformed data—not to reproduce CPCB’s validated pipeline from an uncalibrated household sensor.

A seven-field check before sharing an AQI

  1. Source: Is it CPCB, a government forecast, a research network or a private sensor platform?
  2. Station: Which exact monitor or city aggregate is represented?
  3. Time: When was the last valid update, and in which time zone?
  4. Standard: Does the display use India AQI or another national/private scale?
  5. Pollutant: Which pollutant created the worst sub-index?
  6. Data type: Is the number observed, provisional, corrected or forecast?
  7. Window: Is it a 24-hour running AQI, eight-hour gas value or short-term sensor concentration?

A defensible social post says, for example, “CPCB station X showed India AQI category Y at time Z, with pollutant P prominent,” and links the source. It does not turn one station into a claim about every resident of Delhi or crop away an old timestamp.

Observed data and forecasts answer different questions

A monitor tells you what its instruments observed at a particular location over a defined period. A forecast combines emissions information, weather prediction and atmospheric models to estimate future conditions over an area. Forecasts are valuable for planning, but uncertainty grows with time and depends on the model and inputs.

India’s Ministry of Earth Sciences identifies IMD and IITM systems as sources of operational air-quality information and forecasts. When a forecast and current monitor disagree, label both rather than selecting the scarier number. The observation may reflect a local plume; the model may represent a wider grid or future hour.

Low-cost sensors are useful with limits

Optical particle sensors estimate particle concentrations from light scattering. They can reveal patterns across time and support local investigations, but humidity, particle composition, placement, ageing and calibration can affect readings. A balcony device near cooking exhaust or a dusty road is not automatically representative of a neighbourhood.

Use a household sensor to notice trends, compare indoor interventions or decide when to investigate—not to declare a regulatory violation by itself. Keep it away from direct steam, smoke and strong airflow unless that source is what you intend to measure. Record model, location, maintenance and any calibration applied.

Visibility and smell are unreliable AQI instruments

Haze can result from particles and humidity; clear-looking air can still contain elevated invisible pollutants. Smell detects some compounds at low levels and misses others entirely. Loss of smell through adaptation can make a persistent exposure seem to disappear. A photograph can document visibility but not establish PM2.5, ozone or nitrogen dioxide concentration.

Similarly, one dramatic image should not be used as a live citywide reading. Use timestamped monitor data, weather context and station metadata. When reporting a pollution episode, separate what instruments show from hypotheses about sources unless a credible source-apportionment analysis is available.

What a single AQI cannot tell you

AQI does not calculate an individual dose, which also depends on time, breathing rate, activity, building filtration and movement between microenvironments. It does not identify every chemical or source. It does not prove why a value changed. It does not measure indoor air unless an indoor instrument is involved.

A city average also cannot establish conditions at a school gate, construction site, kitchen or bedroom. The index is a useful warning and communication layer, not a complete exposure assessment. For a practical indoor intervention, RawColumn’s safer DIY box-fan air-cleaner guide explains filter fit, electrical limits and how to avoid overstating performance.

How to use AQI information without overclaiming

Start with the official station view, then look at nearby stations and the forecast. Note whether they agree on category and prominent pollutant. Follow CPCB or local authority public guidance for the current band, and check school, workplace or clinical advice relevant to you. Avoid universal instructions based only on one app tile.

For symptoms such as severe breathlessness, chest pain, confusion or blue lips, seek urgent medical care rather than monitoring an app. Routine masks, exercise changes, ventilation and air cleaning involve fit, building conditions and personal health; obtain clinical guidance when you have asthma, COPD, heart disease, pregnancy or another higher-risk condition.

Five common claims, corrected

“The AQI is 350, so PM2.5 is 350 µg/m³.”

Not necessarily. AQI and concentration are different quantities. Open the pollutant detail to find the measured concentration and unit.

“My app updated one minute ago, so it is a one-minute air sample.”

Not necessarily. The timestamp may mark a refreshed running average that includes many previous hours.

“The city average is lower than my station, so my station is wrong.”

No. An average can sit below a local hotspot and above a cleaner location. Check station status and spatial context.

“One app uses US AQI and one uses India AQI, but the numbers should match.”

No. Breakpoints and conventions differ. Compare pollutant concentrations and averaging periods or select the same AQI standard.

“A severe day proves one specific source caused it.”

No. AQI reports ambient conditions, not source attribution. Source claims need emissions, meteorology and chemical or modelling evidence.

Why measurement literacy matters for science policy

Good public science preserves the chain from instrument to claim: sampling, quality control, averaging, conversion, uncertainty and communication. Removing any link can turn a useful warning into false precision. The same discipline matters in research funding and evaluation; RawColumn’s explainer on India’s Horizon Europe association talks shows why institutions, methods and conditions must be read beyond a headline.

AQI is most useful when treated as a carefully defined index. Share the station and time, name the standard and prominent pollutant, distinguish observation from forecast, and retain uncertainty. That makes the information more actionable without pretending one number describes all of Delhi’s air.

Sources

Trend and scientific-source review completed 9 November 2026. Monitoring networks, methods and advisories can change; verify the station timestamp and current official guidance. This article is general science and public-information literacy, not medical advice.

Tags

  • Delhi AQI today
  • live AQI
  • air quality index
  • PM2.5
  • CPCB AQI
  • air pollution science

Related articles

More Science articles