Science · News
Cooking is the largest local PM1 source at mealtimes in London and Beijing
The percentages are shares of particles released at those hours, not the mix in the air people breathe, and the same campaign recorded a different mix in Delhi.
In this story
- Midday primary PM1 from cooking reached 83 percent in London, and the early evening reached 87 percent, against 61 and 76 percent in Beijing.
- Cooking contributed a tenth or less of the PM1 mass measured on the towers.
- In Delhi, cooking was 8 percent of the daily primary total and traffic with combustion stayed larger.
- London fluxes are from 2021, Beijing fluxes from 2016 and 2017, and brake, tire, and road dust were not in the measurement.
Cooking has become the largest local source of primary PM1 during mealtimes in central London and Beijing, measurements led by the UK Centre for Ecology & Hydrology show.
Ben Langford and colleagues published the work on 23 October 2026 in npj Climate and Atmospheric Science, DOI 10.1038/s41612-026-01583-6. Instruments on towers measured the upward flux of non-refractory PM1, particles smaller than one micrometer, and sorted that flux by chemical signature. That size class is not the ultrafine fraction, which sits below 0.1 micrometers. The London readings come from the BT Tower in summer and winter 2021. Beijing was measured in winter 2016 and summer 2017. Delhi was measured in autumn 2018.
At mealtimes, cooking accounted for as much as 83 percent of the midday primary flux in London and 87 percent in the early evening. In Beijing the midday share reached 61 percent and the early-evening share 76 percent. Traffic-related material, including black carbon, fell below 10 percent of primary PM1 only at some of those hours. Across the day its share was 16 to 38 percent in London and 28 to 46 percent in Beijing. The mass spectrometer does not see brake, tire, or road dust, so the traffic total is larger than the exhaust fraction alone.
On the towers, cooking contributed 5 to 10 percent of PM1 mass. Most of that mass is secondary, formed later in the atmosphere, so a large local emission is not most of what a person inhales. The authors say further cuts in urban PM1 depend on those secondary processes.
Delhi did not follow the London and Beijing pattern. Cooking there averaged 8 percent of daily primary PM1, while traffic and combustion remained the main sources and traffic-related emissions were 56 percent. The absolute cooking flux in Delhi was still in the same range as London. The comparison is between different neighborhoods and seasons, not one citywide inventory.
Langford, an atmospheric scientist at the UK Centre, said most people smell cooking and do not think of it as a pollutant, and that filtration of commercial kitchen fumes could reduce this part of the flux relatively quickly. Co-author Eiko Nemitz said policies aimed at vehicle exhaust have been effective enough that cooking is now a large, poorly regulated part of the mix. Those remarks are from the Centre's 3 November 2026 account of the paper.
The London Atmospheric Emissions Inventory has only recently included cooking, so the study cannot show a historical trend for that source. The winter London sample also overlapped COVID-19 Plan B measures, which the authors say likely reduced restaurant traffic, and the Beijing winter footprint covered a more restaurant-heavy district than the summer one.