Sign in

How a balloon tells us whether Delhi’s pollution has room to spread out

Twice a day, IMD’s balloon from Safdarjung measures the air above Delhi. This week it found the city’s pollution held in a layer barely 400 metres deep.

Published on: Nov 13, 2026, 16:35:32 IST
Share
Share via
  • facebook
  • twitter
  • linkedin
Copy link
  • copy link

For most of this week, Delhi breathed from a shrinking room. On Tuesday afternoon the city’s smoke, dust and exhaust had 1,411 metres of air above the ground to spread through; by Thursday that ceiling had dropped to 396 metres, lower than on a usual afternoon in early January, and PM2.5 across the city more than doubled in a day.

A layer of smog seen over Yamuna, near Kalindi Kunj. (Sunil Ghosh/HT Photo)
A layer of smog seen over Yamuna, near Kalindi Kunj. (Sunil Ghosh/HT Photo)

That ceiling is invisible from the ground, but a weather balloon can find it. The India Meteorological Department (IMD) lets one go from its Safdarjung observatory at about 4.30am and 4.30pm every day, and on its way up it radios back what the air is doing at every height: how warm it is, how far the day's pollution has been stirred upwards, and how hard and from which direction the wind is blowing.

What the balloon measures

The balloon lifts a small instrument pack that reports pressure, temperature, humidity, wind speed and wind direction as it climbs, from the rooftops to several kilometres up. HT keeps only the flights that clear 3,000 metres, and the readings turn up in the public archive about two hours after launch.

The two flights bracket the day. The 4.30am balloon goes up at the end of a long night of cooling, when the air near the ground is at its coldest and stillest; the 4.30pm balloon goes up after the afternoon's strongest stirring, so it can come in a little below the highest point the air reached earlier in the day.

Also read: How CPCB’s defence of Delhi's Diwali AQI undercuts its own case

The dome: How high the air mixes

On a sunny afternoon the sun does most of the work of clearing Delhi's air. As the ground warms through the day, the air touching it warms too and rises in bubbles, hauling pollution upwards and pulling cleaner air down, and the height those bubbles reach is the mixing height. Everything the city emits gets diluted in the layer beneath it, which is why it helps to think of that layer as a dome over Delhi: the taller the dome, the more air there is to thin the pollution out.

Finding its top takes a little physics: rising air expands and cools, by about 10°C for every kilometre, and so HT corrects each of the balloon's readings for that cooling and looks for the first height where the air is 1°C warmer than a bubble rising from the ground would be. Above that line the surrounding air is too warm and too light for the bubbles to push into, and the stirring stops there.

The dome rises and sinks with the seasons. Over the six winters from 2019-20 to 2024-25, the 4.30pm balloon usually found it about 1,800 metres tall in early October, 1,300 metres in November and 1,200 metres in early December, before it sank to about 600 metres in the first half of January, its lowest fortnight, as the days shortened and the sun weakened.

Tuesday, Nov 10, 4.30pm

Every afternoon, the sun warms Delhi's ground and the air above it rises, carrying the city's smoke, dust and exhaust upwards and thinning it out.

On Tuesday that stirring reached 1,411 metres. The layer works like a dome over the city: everything Delhi emits is spread through the air beneath it.

How the balloon finds the top

The India Meteorological Department launches a weather balloon from its Safdarjung observatory at about 4.30am and 4.30pm every day. As it climbs, it radios back the temperature and the wind at every height.

From those readings HT finds the height at which the air becomes too warm and light for air rising from the ground to push into. That is where the afternoon's stirring stopped: on Tuesday, at 1,411 metres.

Wednesday, Nov 11, 4.30pm

A day later, the balloon found the dome less than half as deep, at 676 metres.

Delhi's pollution now had less than half the air to spread through. The usual depth for early November, from six winters of balloon readings, is about 1,300 metres.

Thursday, Nov 12, 4.30pm

On Thursday the dome had shrunk to 396 metres, less than a third of the usual depth for the time of year.

PM2.5 across Delhi rose through the afternoon instead of falling, and by 6pm was 244 micrograms per cubic metre, nearly two and a half times Wednesday's level.

The wind inside the dome

Depth is half of it. The wind blowing through the dome carries pollution away, and that weakened too, from 15 km/h on Tuesday to 8.5 km/h on Thursday.

Multiplied together, the two give the ventilation index, a measure of how much air is available to dilute the city's emissions. It fell from 5,873 to 934 square metres a second, a drop of 84%. IMD treats anything below 6,000 as unfavourable.

Thursday, 4.30am: the wind above

Above the dome the wind was strong, and it came from the north-west, across Punjab and Haryana and the farm-fire belt.

The dawn balloon measured it at 34 km/h between 1 km and 2 km up, from the north-west at every level to 3 km. Two days after Diwali it had been 5 km/h.

Against the season

Over the six winters from 2019-20 to 2024-25, the afternoon dome was usually about 1,800 metres deep in early October and shrank to about 600 metres in early January, its shallowest fortnight.

Thursday's 396 metres was shallower than a usual January afternoon, in the middle of November.

Source: IMD weather balloon, Safdarjung (University of Wyoming archive); CPCB. HT analysis. Dots are illustrative and show an equal amount of pollution in each dome; buildings are illustrative. The orange mark on the height scale is the Qutub Minar, 73 m.

The lid: What happens at night

After sunset the order flips. The ground sheds its heat quickly and chills the air touching it, while the air a few hundred metres up holds on to its warmth, and that warm layer acts as a lid: the cold air underneath, loaded with the evening's smoke and dust, is too heavy to rise through it.

Delhi gets this lid almost every winter night. Across the six winters, the 4.30am balloon found one on 94% of mornings from October to January and on 97% of November mornings; on Thursday it found the air 207 metres up 6.2°C warmer than at the ground, 24.2°C against 18°C.

By mid-morning the sun usually breaks the lid open. On Thursday it did, but only just: the 4.30pm balloon found no lid, yet above about 170 metres the air cooled with height far more slowly than a rising bubble does, so the bubbles quickly ran out of lift and the dome topped out at 396 metres.

The wind inside the dome

A tall dome is only half the cure. Still air in a deep dome leaves pollution to stew, while a strong wind through a shallow one can still sweep pollution away, so HT multiplies the dome's depth by the average wind blowing through it to get the ventilation index, a single number for how much air the city has to dilute its own emissions.

IMD counts a ventilation index below 6,000 square metres a second as unfavourable for clearing pollution. Delhi's was 5,873 on Tuesday afternoon, 2,290 on Wednesday and 934 on Thursday, a collapse of 84% in two days to less than half the usual figure for early January.

Down at street level, Delhi's winter air hardly moves. In October and November 2024, 66% of the reports from IMD's Safdarjung observatory recorded calm air, against 26% at Indira Gandhi International (IGI) airport, 15 km away on open ground, and at night Safdarjung was calm in 79% to 98% of its reports; HT has not yet worked out how much of that gap comes from the city crowding the station and how much from where the instrument stands.

HT Immersive: The smog Delhi can't wish away

The wind above the dome

A kilometre or two up, the wind is far stronger, and it mostly comes from one direction. Over the six winters, the 4.30am balloon found the wind between 1km and 2km blowing from the west, north-west or north on 82% of October mornings and 74% of November mornings, typically at about 21 km/h; from the east, it blew on just 2% to 5% of mornings.

That northwesterly has to cross Punjab and Haryana to reach Delhi, and in October and November its route runs straight over the farm-fire belt; on Thursday morning the balloon clocked it at 34 km/h, from the north-west at every level up to 3 km, two days after it had slumped to 5 km/h. Smoke riding in above the dome can be dragged down to street level once the afternoon stirring reaches it, though the balloon alone cannot say how much of it does.

The wind that changes things

In HT's daily readings this season, a turn of the wind has done as much as rain to clear the air, and sometimes more: on October 11 and 12, most of the fall in Delhi's PM2.5 came with a change of wind on the Sunday evening, hours before 42 mm of rain fell early on Monday.

Rain does scrub particles out of the air, but HT's daily reading credits it only when it is more than a passing shower, 2.5 mm or more at Safdarjung, the amount IMD counts as a rainy day.

Coming up

Two numbers from the 4.30pm balloon tell most of the story each evening: the mixing height and the ventilation index, set against the usual for the fortnight. A dome under about 1,000 metres with less than 10 km/h of wind through it leaves the day's pollution little room to thin out, and the night's lid then holds whatever is left over the city until morning.

The worst of the season is still ahead. On the six-winter record, the usual dome shrinks by half between early December and early January, from about 1,200 metres to about 600, and the usual ventilation index falls from about 4,700 to 2,200; this Thursday, Delhi got a January afternoon in the middle of November.

  • Binayak Dasgupta
    ABOUT THE AUTHOR
    Binayak Dasgupta

    Binayak Dasgupta is a senior editor at Hindustan Times, New Delhi, where he has worked since 2013. He heads the paper's Page 1 team and is part of the editorial leadership overseeing daily news operations. He has previously led the national news desk. His work sits at the intersection of specialist depth and emerging public consequence — stories where complexity and nuance are in themselves the journalistic challenge. He shapes Hindustan Times's coverage and editorial positions on technology and artificial intelligence, information security, environment, health and science — steering these subjects from the commissioning of reporting to editing and the framing of the paper’s institutional view. His work spans analysis, commentary and reportage and in the past has led the paper's coverage of Covid-19 pandemic science. His interviews and hosted sessions include Nobel laureates, technology leaders, economists and climate justice scholars. In recent months, Binayak has overseen the coverage of the Air India Flight 171 crash and its investigation, India's energy exposure during the West Asia conflict, the trade dispute with the US, Delhi's air quality emergency and the evolving conversations around AI, including the IndiaAI Summit.Read More