Skip to content
Environment Dashboard

Data

Deaths from fossil-fuel air pollution (fine particles and ozone), world (Lelieveld et al. 2023)

Excess deaths worldwide in 2019 from breathing outdoor fine particles (PM2.5) and ozone from burning coal, oil and gas, as modelled by Lelieveld and colleagues (BMJ, 2023): a harm of the burning itself, not of the warming. Calculated from the authors' deposited model runs as the deaths in the reference run less those in a run with all fossil-fuel emissions removed, as the paper defines them. The paper's 95% confidence interval is quoted with the value; a range cannot be calculated by subtracting two ranges, so none is calculated here. A twin of the Lancet Countdown's estimate (indicator 3.2.1), from a different model; the two are never combined.

Latest value

5,126,627deaths

World, 2019. Version Edmond dataset version 2 (2023-11-23).

Deaths from fossil-fuel air pollution (fine particles and ozone), world (Lelieveld et al. 2023) (deaths)

1 values from 2019 to 2019.
Show as a table
Deaths from fossil-fuel air pollution (fine particles and ozone), world (Lelieveld et al. 2023), every value
PeriodValue (deaths)Status
20195,126,627final
Calculated by Environment Dashboard from Pozzer, A. (2023). Fossil fuel-attributable air pollution deaths, version 2. Edmond, Max Planck Society. doi:10.17617/3.T7XAXH. CC0 1.0. Data for Lelieveld, J., Haines, A., Burnett, R., Tonne, C., Klingmüller, K., Münzel, T. & Pozzer, A. (2023). Air pollution deaths attributable to fossil fuels: observational and modelling study. BMJ 383, e077784. doi:10.1136/bmj-2023-077784. Changes: the deaths in the run without fossil fuels subtracted from those in the reference run (world, all causes, fine particles and ozone). CC0 1.0.

Notes on individual values

  • Lelieveld et al. 2023 (BMJ 383: e077784), abstract: "An estimated 5.13 million (3.63 to 6.32) excess deaths per year globally are attributable to ambient air pollution from fossil fuel use and therefore could potentially be avoided by phasing out fossil fuels." The range in brackets is the paper's 95% confidence interval.

Where it comes from

Producer
Max Planck Institute for Chemistry (J. Lelieveld, A. Haines, R. Burnett, C. Tonne, K. Klingmüller, T. Münzel and A. Pozzer); deposited by A. Pozzer in Edmond, the Max Planck Society's research data repository · Deaths from outdoor air pollution with and without fossil fuels, 2019 (Lelieveld et al. 2023, data deposit)
Version
Edmond dataset version 2 (2023-11-23)
File
https://edmond.mpg.de/api/access/datafile/236841
Fetched
2026-10-12 (automatically), 17,085 bytes
Fingerprint
sha256 37e2ef88859e5eb51ba822eebf4923fdd69bf94ef1d0f70da4fc05aa9dff76ebETag "d7a8158da797fd01950a313f52144f72" · Last-Modified Sat, 17 May 2025 16:49:20 GMT
Copies
Archived copy (Wayback Machine) Our copy of the exact file
Cite the producer
Pozzer, Andrea, 2023, "Fossil fuel-attributable air pollution deaths", https://doi.org/10.17617/3.T7XAXH, Edmond, V2
Producer
Max Planck Institute for Chemistry (J. Lelieveld, A. Haines, R. Burnett, C. Tonne, K. Klingmüller, T. Münzel and A. Pozzer); deposited by A. Pozzer in Edmond, the Max Planck Society's research data repository · Deaths from outdoor air pollution with and without fossil fuels, 2019 (Lelieveld et al. 2023, data deposit)
Version
Edmond dataset version 2 (2023-11-23)
File
https://edmond.mpg.de/api/access/datafile/236835
Fetched
2026-10-12 (automatically), 17,118 bytes
Fingerprint
sha256 5e1fe7bc2c7417399502c0a93e9d248bfb8a54d46319ee592dc92dc11d016010ETag "3be16ab53e3f0f223a66ea4eb22919a5" · Last-Modified Sat, 17 May 2025 16:49:20 GMT
Copies
No Wayback copy recorded yet.Our copy of the exact file
Cite the producer
Pozzer, Andrea, 2023, "Fossil fuel-attributable air pollution deaths", https://doi.org/10.17617/3.T7XAXH, Edmond, V2

What was done to it

  1. Read the WORLD row of mortality_region_2019_ref.xlsx (the reference model run) and of mortality_region_2019_noFF.xlsx (the run with all fossil-fuel emissions removed) from the Edmond deposit of Lelieveld et al. 2023: column 'NCD+LRI+LRI5-+CPO3', excess deaths in 2019 from all causes due to outdoor fine particles and ozone. Checked that both files list the same regions with the same population, and that the reference total and its 95% confidence limits round to the paper's printed 8.34 million (5.63, 11.19 million).

    Code: pipeline/src/envdash/transforms/impacts/lelieveld_2023.py · code sha256 81c07c7da3da1454… · inputs 37e2ef88859e… 5e1fe7bc2c74…

  2. Subtracted the deaths without fossil fuels from those of the reference run, in exact decimal arithmetic on the stored values: the deaths the paper attributes to fossil-fuel emissions. Confidence limits cannot be subtracted, so no range is calculated; the observation quotes the paper's own 95% confidence interval.

    Code: pipeline/src/envdash/transforms/impacts/lelieveld_2023.py · code sha256 81c07c7da3da1454… · inputs 37e2ef88859e… 5e1fe7bc2c74…

What it covers

Geography
World
Basis
Modelled excess deaths from all causes due to long-term exposure to outdoor fine particles (PM2.5) and ozone in 2019: exposure from the EMAC atmospheric chemistry model (2019 emissions and meteorology) combined with NASA satellite-based fine particles and population, and deaths from Burnett and colleagues' FUSION relative-risk model applied to Global Burden of Disease 2019 baseline mortality. Deaths from fossil fuels are the reference run less a run with every fossil-fuel emission removed at once. The authors give three reasons their estimate is higher than most earlier ones: the new relative-risk model, counting deaths from all causes as well as from specific diseases, and the large fall in exposure when fossil fuels are removed entirely.
Unit
deaths (deaths)