Two-Thirds Of Delhi Does Not Have Reliable Air Quality Data

According to the University of Chicago’s 2025 Air Quality Life Index (AQLI) data, an average resident of the national capital region would gain over eight years of life if the city’s air quality met the World Health Organization guideline of 5 micrograms of PM 2.5 per cubic metre of air (µg/m³).

As smog levels worsened through recent weeks, the health emergency has also produced an unusual public response: small but growing protests led by residents, environmental groups and public-health advocates, calling for accountability, stricter enforcement and clearer communication from authorities.

But there is a need for a robust and reliable system to measure the crisis. Policies, enforcement, and even emergency responses are built on data. And that raises a far less discussed question: Does Delhi actually have the monitoring infrastructure needed to understand its pollution at a granular, neighbourhood-level scale?

Air pollution monitoring, experts say, serves three distinct purposes, and Delhi’s network reflects only part of that design. The first is regulatory monitoring: stations that check whether a city is complying with national air-quality standards.

Delhi currently has 38 Continuous Ambient Air Quality Monitoring Stations (CAAQMS) for this purpose. These are automated, high-precision stations that measure pollutants in real time tracking eight pollutants: PM 2.5, PM 10, nitrogen dioxide, sulphur dioxide, carbon monoxide, ozone, ammonia and lead. They transmit the data directly to central servers with minimal human interference.

In theory, they form the backbone of the city’s regulatory system, generating the numbers that classify days as “good,” “poor” or “severe,” and informing emergency measures during pollution spikes.

“Regulatory monitoring tells you whether the city is meeting national standards or not,” said Chandra Bhushan, CEO of the International Forum for Environment, Sustainability & Technology (iFOREST). CPCB’s norms say a city with over five million people needs 12 continuous monitoring stations for basic regulatory coverage. Delhi, with the current population of 21.8 million, going by scaling the CPCB’s minimum Delhi needs 52 such stations. “Even so,” Bhushan said, “the existing network is at least able to indicate whether the city is within permissible limits or not, and how often.”

But patchy regulatory monitors alone cannot guarantee accuracy. This is where the second layer—sentinel or reference monitoring—comes in. These are lab-grade, high-precision instruments that periodically verify whether the regulatory monitors are functioning correctly. “You need reference monitors to check if your real-time stations are working properly,” Bhushan explained. “They are the quality-control system.”

However, the system begins to thin out at the third and most operationally crucial layer: local-area planning monitoring. “These are smaller, neighbourhood-scale stations meant to detect hyper-local spikes—from construction sites, industrial clusters or traffic bottlenecks—so authorities can respond quickly” says Bhushan.

This is precisely the layer Delhi lacks, experts say. “In the absence of strong local-monitoring stations, the regulatory network ends up doubling as the only government source of neighbourhood-level data—and that’s not what it’s built for,” said Sunil Dahiya, founder and lead analyst at Envirocatalysts, a Delhi-based environmental think tank. “You cannot plan action for specific neighbourhoods if you don’t measure them properly.”

An air quality monitoring (AQI) station typically represents an area covering a 2 km radius, which translates to about 15 sq km. (For context, Delhi is spread across an area of 1,483 sq km). When the location of Delhi’s monitoring stations is plotted on a map, the coverage gaps become immediately visible. The densest cluster of monitors lies in Central, New Delhi, East Delhi, and parts of South Delhi, where the city’s administrative, commercial, and institutional zones are concentrated.

As per IndiaSpend’s analysis only 34.5% of Delhi’s geographical area is covered in a 2-km radius of a monitoring station. Many of these neighbourhoods—such as Civil Lines, Chandni Chowk, Mandir Marg, India Gate, and Lodhi Road—have overlapping 2 km zones, effectively giving them multiple layers of measurement.

But farther from the city’s core, the coverage thins out dramatically.

Large parts of outer Delhi—especially Bawana, Chhattarpur, Aya Nagar, Karawal Nagar, Kakrola, Gokalpur, and the rural-urban fringe in the North and Southwest districts—fall almost entirely outside these 2 km catchments. These are also the areas where informal industries, brick kilns, waste burning, and traffic from inter-state freight movement converge—making monitoring even more critical.

Similarly, large stretches along the Yamuna floodplains, peri-urban pockets on the eastern edge bordering Ghaziabad, and belts in the northwest between Mundka and Tikri show very low or no direct coverage, despite hosting waste-to-energy plants, industrial godowns, and major highways.

With around 0.14 air quality monitors per million Indians, the country’s monitor density is well below that of countries such as China, the United States, Japan and Brazil, as IndiaSpend reported in December 2021. While ramping up its air quality network, India will also have to complement it with satellite images and low-cost sensors to gather adequate data that give the full picture of existing air quality, we had reported.

Why local monitoring matters

Local monitoring can be transformative because it shifts the focus from city-wide averages to what people are actually breathing in their own neighbourhoods. Dahiya explains, “We have enough data to act on major polluting sources in the broader geography,” he said. “But certain locations in cities behave very differently, and the exposure to people there is high. There are multiple pollution hotspots across the city—and to detect them directly, we need data from those hotspots.”

According to Dahiya, a larger and more fine-grained network does more than just improve scientific precision—it makes the crisis personal. “When more data becomes available at the locality level, two things happen,” he said. First, people stop thinking of pollution as a city-wide abstraction—Delhi is polluted—and start recognising what they breathe every day: my neighbourhood is polluted. That shift, he said, increases both the urgency to act and the pressure to demand accountability.

Second, hyperlocal readings can drive behavioural change. Residents learn to avoid early-morning workouts on high-pollution days, reduce time spent outdoors, and connect spikes in PM 2.5 to local triggers—heavy traffic, generators, waste burning—linking cause and effect in real time. “Their activities will change towards low-emitting behaviours once they see the relationship between what’s happening around them and the air they breathe,” he said.

This is also why expanding Delhi’s official network—currently concentrated heavily in the city’s core—is becoming harder to ignore. Bhushan agrees that the system needs to grow, but he notes a practical barrier: each reference-grade continuous monitoring station costs around Rs 1 crore to install and maintain. For a city the size of Delhi, scaling up purely through government-grade monitors is financially unrealistic.

“That’s where low-cost monitors can play a role,” Bhushan said.

This gap has given rise to a parallel ecosystem of low-cost and portable sensors. As Rohit Negi, associate professor of urban studies at Ambedkar University, Delhi, notes in Atmosphere of Collaboration: Air Pollution Science, Politics and Ecopreneurship in Delhi, a new generation of “airpreneurs” has built devices that allow neighbourhoods, residents’ associations and citizen groups to track their “proximate air”, helping create more granular maps of pollution and filling spaces between what he calls the city’s “relatively sparse government reference monitors”.

But low-cost sensors come with significant caveats. Because they are highly sensitive to location, they can produce misleading readings when placed in non-representative sites—for example, directly next to a busy road, a construction site, or used hand-held. “We need low-cost monitors for more granular data,” Dahiya said, “but they must be placed in a way that they get representative air samples” that can be meaningfully compared to government monitors. Other experts emphasise that data from such devices should be used cautiously unless the equipment is properly calibrated and maintained.

Bhushan adds an important distinction: low-cost monitors are not meant to replace official stations, nor to deliver precise regulatory data. Their value lies elsewhere. “They are meant to identify where the problem is,” he said. “Even if they are plus-minus twenty in accuracy, it doesn’t matter—they’re for spotting issues and taking action. They are not supposed to be very precise.”

Still, as both Dahiya and Bhushan point out, a hybrid system—reference-grade stations for regulatory baselines, and a dense network of low-cost monitors for neighbourhood-level diagnostics—may be the only realistic way to close Delhi’s coverage gaps and build the kind of public engagement that forces political action.

In 2015, for instance, IndiaSpend piloted the #Breathe project, deploying low-cost monitors—at its peak—across 28 sites in Delhi, and a total of 110 devices across 14 cities. Using these monitors, we were able to provide real-time analyses: showing how the odd-even measure was not enough to address pollution (a finding confirmed months later by academic studies), which times of the day saw most pollution, and explaining spikes and other factors.

Problem with the data

In late October, videos emerged of water sprinklers operating right next to certain monitoring stations—a practice that can artificially suppress readings in the immediate surroundings. Then came findings from a Newslaundry investigation: of 25 real-time air-quality monitoring stations they surveyed across the capital, at least 22 were found to be placed in violation of siting guidelines. The norms are explicit: no trees within 20 metres, and the distance between a monitor and any obstacle must be twice the obstacle’s height above the sampling inlet. Yet several stations were squeezed between buildings, shaded by trees, or located beside busy intersections, conditions that compromise data accuracy.

Operational reliability has also been shaky. Several reports have highlighted Delhi's pollution monitoring app failed during a severe air quality crisis in the past one month. This is similar to the situation a decade ago, when IndiaSpend had reported that 11 monitors in the city were non-functional for half the time.

A few missing hours can be written off as routine calibration or technical glitches, but prolonged blackouts signal deeper problems. Dahiya said that when stations fail for many hours or days, it points to a “fundamental issue with the sensor” and warrants an audit and upgrades. Another reason for the Diwali night blackout, he noted, could be sensor overload. When pollution spikes beyond what an instrument is designed to measure, it can freeze or shut down entirely. “That can happen,” he explained, “but only if every sensor stops reporting after crossing a particular threshold.”

What Delhi needs now is not just more stations, but a monitoring architecture that is wider, more reliable and more locally responsive—one that combines robust reference-grade stations with a dense layer of properly placed, properly maintained low-cost sensors.

IndiaSpend has reached out to the CPCB and the Delhi Pollution Control Committee for their comments on future expansion plans. We will update this story when we receive a response.

We welcome feedback. Please write to respond@indiaspend.org. We reserve the right to edit responses for language and grammar.