Cause 03 · Peaks in summer
Road and construction dust
Direct answer: dust from roads, building sites and open land makes coarse particles (PM10). It is the biggest source of PM10 in Delhi and becomes the top problem in dry summer months.
At a glance
- Dust (soil, road and construction) gave about 17% of Delhi's winter PM2.5 in the TERI–ARAI study, and a larger share in summer.34,36
- About two-thirds of the money spent under the National Clean Air Programme has gone to dust control.86
- Every construction site of 500 m² or more in Delhi must register on a dust portal and file self-audits.22
- Our model analysis: coarse dust was about 76% of PM10 in May but only 4% in January.1
How dust pollutes
Traffic lifts fine dust from road surfaces and unpaved edges back into the air. Building sites, demolition, open trucks carrying sand and debris, and bare land add more. In summer, hot dry winds also bring dust from Rajasthan and beyond. Most dust particles are coarse (PM10), so they settle faster than smoke, but they still reach the lungs and add to the AQI.
What research says
- TERI–ARAI (2018): dust from soil, roads and construction gave about 17% of Delhi's winter PM2.5. Across the study's methods and seasons the dust share ranged from about 16% to 38%.34,36
- Where the money goes: an analysis of National Clean Air Programme spending found about two-thirds went to dust management and less than 1% to industrial pollution.86
- Evidence gap: sprinklers, anti-smog guns and sweepers are widely used, but we did not find any published citywide study of how much they lower PM10.
What our data shows
Show the numbers
| Month | Coarse share of PM10 |
|---|---|
| January | 4% |
| April | 78% |
| May | 76% |
| September | 28% |
| 1–10 November | 37% |
Over the full year, the model shows two different kinds of bad air. In April–May, coarse dust is about three-quarters of PM10, and the air mixes up to about 3 km high. In winter, PM2.5 is highest and dust is a smaller part.1,9 293 construction sites are mapped in OpenStreetMap in the study area.11
| Month (2026) | PM2.5 | PM10 | Coarse share of PM10 | Ventilation index (m²/s) |
|---|---|---|---|---|
| January | 121 | 125 | 4% | 2,754 |
| April | 73 | 401 | 78% | 14,345 |
| May | 103 | 486 | 76% | 15,209 |
| September | 52 | 93 | 28% | 4,846 |
| 1–10 November | 115 | 187 | 37% | 4,199 |
What is being done
Large construction sites in Delhi must install PM sensors and video monitoring, and the biggest must run anti-smog guns.22 For winter 2025-26 the city listed 86 mechanical road sweepers and 362 anti-smog guns.88 See Solutions for the next steps we recommend.
Limits of this data
The CAMS model's coarse particles are almost entirely desert dust (rank correlation 0.998). It does not include local road or building dust, so summer values are mostly dust storms. Delhi needs more ground PM10 monitors to measure local dust.