Delhi’s air turned “severe”, the worst grade on the government’s pollution scale, at 6am on Friday on the same 39 monitoring stations the city had last Diwali, an HT analysis of Central Pollution Control Board (CPCB) data shows.
Safdarjung reported dead calm through most of Thursday night. The count usually quoted, which Nasa publishes, comes from a satellite that passes over Punjab once in daylight, at about 2pm. By that count this has been a light season. The two satellites detect fires in different ways, so their growth can be compared but their totals cannot. Nasa’s count is accurate for what it measures, fires burning at about 2pm. Fires lit after the satellite has passed don’t show up in it, and they may now be a large and growing share of the burning. There has been no wind at ground level to clear the air, and pre-Diwali traffic has jammed the city’s roads for days. Every morning this week, a layer of warmer air a couple of hundred metres up has worked like a lid, keeping the night’s pollution close to the ground. Exhaust from the traffic, and the rest of Delhi’s own emissions, build up in that shallow layer. The data do not show what share of Delhi’s pollution comes from farm fires. Nasa counts cover Punjab and Haryana, nominal and high confidence, with fixed hot spots such as factories left out. Winds come from the India Meteorological Department’s balloon soundings and airport reports at Safdarjung.
It grew by only about 60% over the same days. It recorded 184 fires a day in the week to Thursday, close to last year’s 200, half of 2024’s 388, and far below the several thousand a day of 2018-2022. On a typical early-November afternoon, the sun breaks that lid and spreads pollution through about 1,300 metres of air. This week the pollution spread only 600 to 800 metres up, so the same pollution was packed into about half the usual space.
Because its counts for the same dates in 2025 run far higher for reasons not yet explained, iNSAT counts are compared only within this season,.
