Did Diwali Worsen the Air?
AQI in 262 Indian Cities, 3–13 November 2026
A simple before-and-after look at CPCB’s daily city AQI around Diwali (Sunday 8 November 2026): compare each city’s air on Diwali night with the days before and after.
Yes. AQI rose sharply on Diwali night, then fell back.
Average AQI across the 262 cities went from 132 before the festival (3–6 Nov) to 193 on Diwali night (9 Nov). The jump from 8 to 9 Nov (+36 points) was about four times the usual daily rise before the festival (+8), and almost all of it reversed the next day (−33). Air was already getting worse through early November, so not all of the rise is due to Diwali, but the one-day spike points clearly to it.
City AQI
CPCB’s Air Quality Index for each city at 4 PM, averaged over the previous 24 hours. For cities with several monitoring stations it is the average across stations. A higher AQI means worse air.
Why 9 November is “Diwali night”
Diwali (Lakshmi Puja) was on the evening of Sunday 8 Nov 2026. Because each day’s AQI covers the 24 hours up to 4 PM, fireworks on the night of 8 Nov appear in the 9 Nov value (8 Nov 4 PM – 9 Nov 4 PM).
Baseline and after periods
Baseline = a city’s average AQI over 3–6 Nov, before the festival evenings began. After = its average over 10–13 Nov. 7–8 Nov (Dhanteras and Choti Diwali evenings) are shown but kept out of the baseline.
Change vs baseline
How much worse (or better) a city’s air was on Diwali night than in its normal pre-festival days.
Change = AQI on 9 Nov − baseline average% change = change ÷ baseline × 100
Day-on-day jump and peak day
Jump = AQI on 9 Nov − AQI on 8 Nov. Peak day = the date with a city’s highest AQI between 3 and 13 Nov. The daily trend uses only the 215 cities that reported every day, so changes are not caused by cities dropping in or out.
Which cities were kept
A city is analysed if it has a 9 Nov value, at least 3 of the 4 baseline days and at least 3 of the 4 after days. 262 of 299 cities qualify. 36 were dropped for missing data, and Kalyan was dropped because the source lists it twice each day with different values. See the “Cities excluded” tab.
A clear one-day spike
Average AQI jumped 36 points from 8 to 9 Nov, the largest daily change in the period, and fell 33 points the next day. A spike that appears and disappears this fast matches the timing of Diwali night.
Most cities were affected
228 of 262 cities (87%) had worse air on 9 Nov than at baseline, with a median rise of 50%. 125 cities hit their worst day of the period on 9 Nov; no other day had more than 24.
Small baselines saw the biggest jumps
Cleaner cities saw the steepest relative rises: Bettiah went from 34 to 304, Puducherry from 38 to 259 and Dehradun from 56 to 257. Already heavily polluted cities such as Delhi (370 → 344) did not rise further.
Air was getting worse anyway
Average AQI rose about 8 points a day before the festival, and the 10–13 Nov average (151) is still 15% above baseline. Seasonal factors were pushing AQI up, and Diwali added a sharp spike on top.
- Timing evidence, not proof of cause. The data show the spike lines up with Diwali night. They cannot rule out weather (calm winds, low temperatures), crop burning or other sources on the same day; there is no weather or emissions data here.
- Rising seasonal trend. AQI was climbing before the festival, so part of the gap between baseline and 9 Nov would probably have happened without Diwali. The day-on-day jump (+36 vs +8 a day before) is the cleaner measure of the festival effect.
- Daily averages hide the night-time peak. Each value is a 24-hour average, so the hourly peak on Diwali night was likely much higher than these figures show.
- One year, one window. There is no comparison with non-festival years or the same dates in another year.
- Regional timing. In some states (e.g. Tamil Nadu) the main celebration is on the morning of Naraka Chaturdashi. The cities’ states are not in the data, so all cities are treated alike.
- Station coverage. Some cities report from only part of their monitoring network (station coverage is shown in the City Explorer).