Delhi’s pollution-monitoring system produced conflicting accounts of the city’s worst air on Diwali, with hourly data from dozens of stations disappearing for hours just as particulate concentrations reached extreme levels. The episode raises a question more consequential than the precise peak reading: can residents, hospitals and administrators rely on the city’s official air-quality picture during the moments when accurate information matters most?
The immediate discrepancy involved the Central Pollution Control Board’s (CPCB) public data and readings displayed by the Delhi Pollution Control Committee (DPCC). The Delhi University North Campus station recorded hourly PM2.5 concentrations of 999 micrograms per cubic metre at 11 pm on Diwali, after which its real-time hourly data was unavailable for six hours. CPCB’s website resumed displaying data from 5 am on Monday, when the concentration shown had fallen to 723 micrograms per cubic metre.
The interruption was not limited to one station. Delhi has 60 continuous ambient air-quality monitoring stations, operated by the DPCC, CPCB, India Meteorological Department and the Indian Institute of Tropical Meteorology in Pune. According to data released by CPCB, hourly information from at least 35 stations was initially unavailable. The data gap occurred during the overnight period when pollution from fireworks and other sources was reported to have reached its highest levels.
The episode also exposed a technical weakness in how extreme pollution is represented. Several stations stopped releasing readings after reaching PM2.5 thresholds between 800 and 1,000 micrograms per cubic metre, according to the report. India’s AQI is capped at 500, but there is no corresponding upper limit for pollutants such as PM2.5 and PM10. This creates two separate measurement problems: the public index cannot communicate how far beyond the hazardous range pollution has risen, while the monitoring system may not consistently display the underlying pollutant readings at precisely those levels.
That distinction matters because an AQI of 500 does not mean that all conditions above that level are equivalent. The cap makes the index easier to communicate, but it can also flatten very different exposure conditions into the same maximum category. If the hourly pollutant readings used to calculate the index are missing at the same time, the resulting citywide number may provide an incomplete account of the episode rather than simply a less detailed one.
The DPCC’s data introduced a second layer of uncertainty. Jahangirpuri was initially shown with a peak hourly PM2.5 reading of 719 micrograms per cubic metre at 1 am on Monday. The figure was later changed to 4,985 micrograms per cubic metre, reportedly recorded three times, before being revised again to 776 micrograms per cubic metre at Diwali midnight. These are not minor variations. They produce radically different descriptions of the same monitoring period and make it difficult for residents and researchers to determine which number should be treated as authoritative.
The contrast between the two public data systems was equally stark. CPCB data showed the highest hourly PM2.5 level at 1,000 micrograms per cubic metre at Siri Fort and indicated that no station crossed that mark. DPCC data, however, showed 11 locations recording hourly concentrations above 2,000 micrograms per cubic metre, including four above 3,000. Mundka’s hourly PM2.5 level was shown as 3,744 micrograms per cubic metre at 2 am on Monday, while CPCB did not share Mundka readings from midnight to 4 am.
These differences affect more than the historical record. CPCB compiles data from the monitoring network and calculates the city’s AQI. Delhi’s AQI on Monday was calculated using data from 47 of the 60 stations, with New Moti Bagh identified as defunct. The report states that the city would have recorded a higher AQI if data had not been missing for several hours. In other words, the selection and availability of station readings can influence the number presented as the citywide condition.
Sunil Dahiya, founder of the think tank Envirocatalyst, described the pattern as a serious reporting gap. He said 57 of 59 stations were reporting readings at 10 pm, but only 23 were reporting at 2 am. Dahiya also pointed to a near-zero reading of 0.1 from Sirifort being included in the city average while other stations were absent from the public data. His concern was that official numbers could understate the air Delhi residents were actually breathing.
Manoj Kumar, an analyst at the Centre for Research on Energy and Clean Air, said missing readings during severe pollution episodes could affect hourly and daily averages and influence the reported citywide AQI. His explanation identifies the central statistical problem: when observations disappear unevenly, the available sample may no longer represent the city’s pollution distribution. A low reading that remains in the dataset can carry disproportionate weight if high readings from other locations are missing.
This is not simply a question of whether one station malfunctioned. The monitoring network is distributed across a large and varied city, where pollution levels can differ substantially between locations. A station in an industrial or traffic-intensive area may capture conditions that are not visible at a relatively cleaner site. If stations go offline selectively during peak pollution, the citywide average may describe the stations still transmitting rather than the full range of exposure across Delhi.
The institutional structure adds complexity. Four organisations operate the city’s continuous monitoring stations, while CPCB shares the data and calculates the AQI after compiling readings. This arrangement can provide a broad network, but it also creates a need for clear responsibility when numbers conflict. The supplied report does not establish whether the discrepancies arose from sensor limits, data-transfer failures, display problems, differing validation rules or another technical cause.
The DPCC’s explanation was that its entire data was available on its website and that an earlier display problem resulted from a technical glitch. A DPCC official said no station had been switched off and that all data was available in the log. CPCB did not respond to the report. That distinction between data existing in an internal log and data being publicly visible in real time is important for a public warning system. During a pollution emergency, delayed access can affect decisions even if the underlying readings are later recovered.
The report also says that data went missing during the previous Diwali. That makes the latest episode more than an isolated display failure, although the supplied material does not establish whether the two incidents had the same technical cause. Repeated disruption during a predictable annual pollution peak would indicate a system that has not been adequately stress-tested for the period when monitoring demand is highest.
For Delhi’s residents, the consequences are practical. AQI information shapes decisions about outdoor activity, school operations, commuting, work and medical precautions. It also informs public communication by authorities. When the index is capped and the underlying readings are incomplete or revised sharply, residents cannot easily distinguish between a very severe episode and an even more extreme one. The problem is not only public anxiety; it is the reduced usefulness of information intended to guide behaviour.
The episode therefore places the emphasis on data governance as much as on pollution control. A functioning monitoring system requires continuous readings, visible timestamps, transparent treatment of missing observations and a clear explanation when values are revised. The supplied report does not say that these standards are absent from the formal system, but the conflicting public records show that they were not sufficiently clear during this episode.
The evidence confirms three points. First, large portions of the monitoring network were not visible on CPCB’s public data during several hours of extreme pollution. Second, DPCC and CPCB displayed materially different peak readings, including at Jahangirpuri and Mundka. Third, the citywide AQI was calculated from an incomplete set of stations. What remains unresolved is why the systems diverged, which readings should be treated as final, and whether the monitoring architecture can reliably capture future pollution peaks. Those questions will determine whether Delhi’s air-quality numbers function as an emergency information system or merely as a delayed record of conditions after the most dangerous hours have passed.
