Delhi's cleanest October in eight years: what the AQI of 166 really means, and why winter will test it

The headline number, in context

Delhi recorded an average Air Quality Index of 166 in October 2026, the Union Environment Ministry said on 31 October, describing it as the city's lowest October average in eight years. The AQI is a 0-to-500 scale on which readings up to 100 count as satisfactory and 101 to 200 count as moderate; above 200, air enters the poor category and worse. October 2026 therefore sits in the moderate band, a clear step down from recent Octobers, but it is not clean air.

The ministry's year-by-year series makes the scale of the change plain. October averaged 223 in 2025, 234 in 2024, 219 in 2023, 210 in 2022, 173 in 2021, 265 in 2020, 234 in 2019 and 269 in 2018. Only two Octobers since 2018 — 2021 and now 2026 — have averaged below 200 at all. Against October 2025, the 2026 figure is roughly 26 per cent lower.

The improvement is not confined to October alone, though it is much smaller over the longer window. For January to October 2026, Delhi's average AQI stood at 167, compared with 170 in the same period of 2025, 184 in 2024 and 201 in 2018. The one year in the series that beats 2026 over those ten months is 2020, at 156, when pandemic restrictions suppressed activity across the city.

The count of better days tells a similar story. Between 1 January and 31 October 2026, Delhi recorded 220 days on which the daily average AQI stayed below 200, the threshold for the good-to-moderate band. That compares with 200 such days in 2025, 201 in 2024, 206 in 2023 and 157 in 2018 — and, again, only 2020 did better, with 224. Days in the poor-to-severe range, above 200, fell to 84 from 104 in both 2025 and 2024, though 2020 again recorded fewer, at 81.

Why the air improved: rain first, everything else second

The researchers who have examined the October figures credit the weather first. Safdarjung, Delhi's main weather station, recorded 127.5 mm of rain in October against a normal of 15.1 mm, according to India Meteorological Department figures. That is more than eight times the month's usual rainfall, and it arrived across a month in which Delhi's air would normally be deteriorating.

Anumita Roychowdhury of the Centre for Science and Environment told The Times of India that the rain washed dust and smoke out of the air, helped pollutants disperse, and cut the smoke drifting into Delhi from farm fires in neighbouring states. Sunil Dahiya, founder of the think tank Envirocatalyst, put it in similar terms: measures across sectors 'might have led to a reduction', but 'the major factor is driven by meteorology'. Neither expert credited any single policy or measure as the decisive cause, and this article does not either.

The rain explanation has one important check on it. Last October was wet too: Safdarjung recorded 89.3 mm of rain in October 2025, about six times the normal, yet that month still averaged 223. Rain alone, in other words, does not explain the full gap between the two Octobers. Something else differed as well — and the biggest visible difference is the calendar, covered in the next section — alongside whatever contribution sector-wise measures made, which the experts quoted describe only in tentative terms.

What the rain did do is buy Delhi a gentler October. Repeated showers scrub particles from the air and break up the stagnant conditions that let pollution accumulate, and October 2026 had far more of them than usual. But rain is not a policy: it cannot be repeated on demand, and its effect ends when the showers stop. That is why the experts quoted are careful to call the improvement largely meteorology-driven rather than a structural turnaround.

The Diwali effect: this year's smoke is not in these numbers

The second major reason October looks so good is timing. Diwali fell on 20 October in 2025, so firecracker smoke was inside last year's October average. In 2026 the festival falls on 8 November, so none of that smoke appears in October's figures at all. Any year-on-year comparison of the two Octobers therefore flatters 2026, and any honest reading of the record has to say so.

An independent analysis of daily pollution data, reported by The Infra Post, tested exactly this point. It found that October 2026 still recorded the lowest average AQI even after excluding Diwali and the following three days from the comparison — but that the improvement over 2025 was smaller once that adjustment was made. The record survives the calendar correction; its size shrinks.

The practical consequence is that Delhi's real test for 2026 has not happened yet. The Diwali week's smoke, which in past years has driven some of the season's worst readings, will land in November's average instead. When November's figures arrive, they — not October's — will show how much of this year's improvement survives the festival and the onset of winter.

There is a further timing point. The Times of India, using Central Pollution Control Board figures going back to 2015, called October 2026 the lowest October on record, while the ministry's eight-year framing calls it the lowest since 2018. The two claims differ only in how far back the comparison runs; on the ministry's own eight-year series, which this article uses throughout, the record stands either way.

Farm fires: a split picture, and the peak is not over

Farm fires are the seasonal source most closely watched every autumn, and the 2026 picture so far is split by state. Satellite monitoring bulletins covering 15 September to 31 October detected 3,071 farm fires across seven states and union territories. Punjab accounted for 1,339 of them, followed by Rajasthan at 638, Madhya Pradesh at 538, Uttar Pradesh at 345, Uttarakhand at 174, Haryana at 35 and Delhi at two.

Haryana's figure is genuinely striking. Its 35 satellite-detected paddy residue-burning events are the lowest for the corresponding period in seven years, down 65 per cent from 99 last year and down from 2,914 in the same period of 2021. Gurugram and seven other Haryana districts recorded no farm fires at all in the period. Programme manager Jinitha Varghese of EnviroCatalysts credited farmer engagement, machinery support, financial incentives and stronger enforcement in Haryana for the decline.

Punjab, however, moved the other way on the daily count that matters most for Delhi. On 31 October alone, the monitoring bulletin detected 292 residue-burning events in Punjab, against six in Haryana, 24 in Rajasthan and ten in Madhya Pradesh. A separate count obtained through a right-to-information response from the Commission for Air Quality Management put stubble-burning cases between 15 September and 29 October at 1,383 across Punjab, Haryana, Uttar Pradesh and Delhi, with Punjab accounting for 1,007. The two counts use different methods and windows, so they should not be added together — but both point the same way: Punjab remains the dominant source.

The season is also not finished. The monitoring window runs to 30 November, and Varghese cautioned that the next two to three weeks would be crucial, as peak harvesting and the short window before wheat sowing could push incidents up. The bulletin itself cautioned that cloud cover could have hidden some fires in parts of Uttar Pradesh, Madhya Pradesh and Uttarakhand. October's farm-fire contribution, in other words, is an interim picture, and the weeks that most affect Delhi's winter air are still running.

What did not improve: 166 is still polluted air

A monthly average of 166 should not be mistaken for safe air. The fine particles known as PM2.5, which penetrate deep into the lungs and pose particular risks to children, older people and those with respiratory conditions, averaged 70.5 micrograms per cubic metre in October up to 30 October, according to The Times of India. India's national standard allows 60 micrograms per cubic metre over a single day — so even this record month's average sat above the daily limit.

Over the longer January-to-October window the picture is mixed rather than uniformly better. Average PM2.5 between 1 January and 29 October fell to 69 micrograms per cubic metre from 72 in the same period of 2025, the ministry reported — an improvement, but a modest one. Over the same window, average PM10, the larger inhalable particles associated heavily with dust, actually rose, to 171 micrograms per cubic metre from 166 a year earlier. Dust, in other words, remains an unsolved part of Delhi's problem even in a good month.

This section is general information about air quality, not medical advice. Anyone with a respiratory condition should follow their own doctor's guidance rather than treating a monthly average as a personal safety signal. A city-wide average also hides wide variation: on any given day, some monitoring stations record far worse air than others, and a moderate month can still contain poor days.

The honest summary is therefore twofold. Delhi's October was genuinely its best in eight years on the ministry's measure, and the improvement survives even a Diwali-adjusted comparison. But the air remained polluted by every health-relevant measure, one major pollutant got worse, and the seasonal forces that normally drive Delhi's worst air had not yet arrived.

The winter wall: why November to February is a different problem

Delhi's air normally collapses in winter for reasons that have little to do with October's weather. As temperatures fall, a layer of warm air can trap cooler, polluted air near the ground, winds weaken, and the same emissions that disperse in October accumulate instead. Crop-residue smoke, vehicle exhaust, dust, industrial emissions and, this year, Diwali firecrackers all arrive into that stagnant air together.

The Delhi government's own winter direction puts numbers on the pattern. It records that in the past three winters, from 1 November to 15 February, Delhi's median AQI was 312 to 342 — roughly twice October 2026's average, and deep into the very poor and severe bands. That is the baseline the current measures are written against, and it shows why an October average, however welcome, cannot be read as a winter forecast.

Two sets of curbs are already in force. Under the Delhi government's winter direction, which applies from 1 November, at least half of office staff must work from home, vehicles registered outside Delhi that do not meet the BS-VI emission standard may not be driven in the city unless they run on CNG or electricity or are emergency vehicles, and dust-raising demolition and outdoor building work stops except on essential public projects. These apply until 31 January, and parking charges double until 28 February.

Separately, the Commission for Air Quality Management ordered on 30 October that government and private offices across Delhi-NCR allow only 50 per cent of employees to work from office, and paused demolition and outdoor construction from 5 to 20 November in Delhi, Gurugram, Faridabad, Sonipat, Ghaziabad and Noida. Railway and metro works, airport projects, inter-state bus terminals, hospitals, water supply and sanitation projects and national-security and defence work are exempt, as are industrial establishments and essential services where working from home is not practical. The commission described the measures as a precaution against the rise in pollution the region typically experiences in winter.

Enforcement, however, already shows a gap between the written rule and the stated approach. Delhi's environment minister, Manjinder Singh Sirsa, told The Hindu that the government would 'mostly request people to follow the directions till Deepavali' and try not to inconvenience people until then — while the direction itself allows prosecution from 1 November, and no written order relaxing the rules had been found. Whether the curbs bite before or only after Diwali is therefore an open question, and Diwali week itself will be the first real test.

What would prove this is real progress

October 2026 will only look like a turning point in retrospect, and several specific readings would have to line up before that claim is justified. The first is November's average itself: if it stays well below the recent November pattern despite Diwali smoke and the onset of winter stagnation, that would be evidence of something more durable than a rainy month.

The second is the fine-particle trend. A sustained fall in PM2.5 across November and December, rather than a single month's dip, and a reversal of the PM10 rise, would show that the improvement reaches the pollutants that matter most for health. A third is the farm-fire peak: if the next two to three weeks pass without the usual late-season surge, particularly in Punjab, one of the winter's biggest variable sources will have stayed contained.

A fourth test is like-for-like comparison. The cleanest future comparison will be a November-to-November one in which Diwali falls in the same month in both years, and a full winter-season comparison from November to February against the 312-to-342 median range of recent winters. Until those comparisons exist, October's record should be reported with its two great qualifications attached: exceptional rain, and a festival calendar that moved the smoke into November.

None of this diminishes what October achieved. A city that has averaged above 200 in most recent Octobers, and 269 as recently as 2018, recording 166 is a genuine improvement in people's lived experience for a month. But Delhi's air problem has always been a winter problem, and the winter of 2026 has barely begun. The honest headline is that Delhi had its cleanest October in eight years — and that the verdict on whether anything fundamental changed will be written in November, December and January.

Editorial disclosure

This article was researched with AI assistance and reviewed by the publisher before publication. Report factual errors through our contact page.