Loading…
PUBLISHED MODELS, CITED MEASUREMENTS
Français English
DATA STATUS
417 series loaded · last import on 2026-10-25

Air quality Fine particles, nitrogen dioxide, ozone and desert dust, against the WHO guidelines

WHO guideline levels (2021): PM2.5 15 µg/m³ over 24 h and 5 µg/m³ as an annual mean; PM10 45 and 15; NO₂ 25 and 10; ozone 100 µg/m³ over 8 h. Bold red: above the 24-hour level. These are health levels, stricter than the European regulatory limits (PM2.5: 25 µg/m³ a year until 2030).
Fine particles PM2.5, last day
Daily mean PM2.5 at the centre of each city, CAMS models (Europe: regional ensemble at 10 km; elsewhere: global analyses at 45 km).

Last day, in µg/m³

CityDayPM2,5PM10NO₂Ozone (hourly max.)Desert dust
Paris 25/10/2026 4 10 7 73 0
Lyon 25/10/2026 5 8 12 62 0
Marseille 25/10/2026 5 9 11 72 0
Lille 25/10/2026 3 8 3 75 0
London 25/10/2026 3 8 8 79 0
Berlin 25/10/2026 3 4 6 68 0
Madrid 25/10/2026 10 16 36 69 0
Milan 25/10/2026 18 23 28 72 0
Delhi (global model) 25/10/2026 51 — — — 154
Beijing (global model) 25/10/2026 88 — — — 15
Lagos (global model) 25/10/2026 19 — — — 16
São Paulo (global model) 25/10/2026 9 — — — 0
Los Angeles (global model) 25/10/2026 18 — — — 0
Cairo (global model) 25/10/2026 22 — — — 17
UTC daily means of hourly values; ozone: highest hourly value of the day, to be compared with care with the 8-hour guideline (the hourly maximum is somewhat higher). Red: above the 24-hour guideline.

European cities: mean PM2.5 per year (µg/m³) and days above 15 µg/m³

City2013201420152016201720182019202020212022202320242025Trend (µg/m³ per year)
Paris 18
180		 14
111		 14
109		 14
110		 13
95		 12
89		 12
87		 10
52		 12
86		 11
73		 10
52		 11
73		 11
76		 -0.44 ± 0.20
-2.9 % per year
Lyon 15
132		 11
64		 16
160		 14
105		 13
94		 10
63		 11
73		 11
70		 11
87		 11
69		 11
54		 12
78		 11
83		 -0.28 ± 0.24
-2.0 % per year
Marseille 16
170		 11
52		 14
148		 14
112		 13
110		 11
56		 11
69		 9
45		 10
52		 11
53		 10
54		 8
37		 8
36		 -0.48 ± 0.20
-3.4 % per year
Lille 17
166		 13
109		 14
109		 15
130		 14
107		 14
118		 13
102		 11
76		 12
93		 12
78		 11
72		 10
67		 11
83		 -0.44 ± 0.14
-2.8 % per year
London 14
111		 13
91		 11
59		 13
95		 12
69		 11
68		 12
74		 10
52		 10
56		 10
43		 9
40		 10
57		 10
62		 -0.33 ± 0.12
-2.6 % per year
Berlin 12
93		 14
134		 14
126		 15
152		 15
124		 14
129		 13
101		 10
70		 12
83		 10
60		 10
51		 9
51		 10
63		 -0.44 ± 0.26
-3.0 % per year
Madrid 12
68		 10
49		 12
87		 12
103		 12
108		 10
65		 11
77		 10
55		 11
49		 10
30		 10
42		 11
85		 10
52		 -0.13 ± 0.10
-1.1 % per year
Milan 30
282		 22
227		 27
286		 26
281		 24
246		 23
233		 25
262		 24
217		 27
240		 21
205		 19
178		 28
189		 28
223		 -0.13 ± 0.48
-0.5 % per year
First line: annual mean; second: number of days above the 24-hour guideline (WHO tolerates 3 or 4 a year). Years with at least 300 days. Trend: least-squares line on the annual means, ± 2 standard errors corrected for autocorrelation.
PM2.5 in four European cities since 2013 (monthly means, µg/m³) Air quality of European cities, CAMS European ensemble via Open-Meteo (daily from hourly) · CC BY 4.0 (Open-Meteo); contains modified Copernicus Atmosphere Monitoring Service information · 10/25/26 Current · provisional
Show the data as a table
Winter maxima (heating, temperature inversions that trap pollution) and spring episodes (fertiliser spreading: ammonium nitrate). Milan, in the Po valley closed by the Alps, remains the most polluted of the European cities followed.
PM2.5 in four large cities worldwide since 2022 (monthly means, µg/m³) Air quality of large cities worldwide, CAMS global via Open-Meteo (daily from hourly) · CC BY 4.0 (Open-Meteo); contains modified Copernicus Atmosphere Monitoring Service information · 10/25/26 Current · provisional
Show the data as a table
Delhi and Beijing are far above the guideline for much of the year; in Delhi the November peak comes from crop burning in Punjab and winter. Coarser global model: urban peaks are underestimated.
Saharan dust in Madrid and Marseille since 2018 (µg/m³, daily means) Air quality of European cities, CAMS European ensemble via Open-Meteo (daily from hourly) · CC BY 4.0 (Open-Meteo); contains modified Copernicus Atmosphere Monitoring Service information · 10/25/26 Current · provisional
Show the data as a table
Desert part of the particles according to CAMS: short, intense episodes when a southerly flow brings Saharan air, sometimes above 100 µg/m³ (“orange sky” of March 2022).
Method and sources

Surface concentrations of the Copernicus Atmosphere Monitoring Service (CAMS), read through the Open-Meteo air-quality API at the centre of each city: for Europe, the regional ensemble of models with data assimilation (Marécal et al., 2015), about 10 km, since 2013; elsewhere, the global analyses (in the line of the reanalysis of Inness et al., 2019), about 45 km, since August 2022 only. UTC daily means of hourly values (at least 18 hours); ozone: hourly maximum. Compared with the WHO guideline levels (2021).

Reading these data
These are models, not stations: they smooth the peaks of busy streets and depend on emission inventories. The two domains (European and global) are not directly comparable. The European series moves from the validated reanalysis to recent analyses: small breaks are possible, and the trend must be read with its error bars.
SOURCES OF THIS PAGE
WHO (2021). WHO global air quality guidelines: particulate matter (PM2.5 and PM10), ozone, nitrogen dioxide, sulfur dioxide and carbon monoxide. Geneva: WHO. ISBN 978-92-4-003422-8.
WHO (2021). WHO global air quality guidelines: particulate matter (PM2.5 and PM10), ozone, nitrogen dioxide, sulfur dioxide and carbon monoxide. Geneva: WHO. ISBN 978-92-4-003422-8.
Licence : Published literature; cited
Marécal et al. (2015)
Marécal, V., Peuch, V.-H., Andersson, C., Andersson, S., et al. (2015). A regional air quality forecasting system over Europe: the MACC-II daily ensemble production. Geoscientific Model Development, 8, 2777–2813. · doi:10.5194/gmd-8-2777-2015
Licence : Published literature; cited
Inness et al. (2019)
Inness, A., Ades, M., Agustí-Panareda, A., Barré, J., et al. (2019). The CAMS reanalysis of atmospheric composition. Atmospheric Chemistry and Physics, 19, 3515–3556. · doi:10.5194/acp-19-3515-2019
Licence : Published literature; cited
Air quality of European cities, CAMS European ensemble via Open-Meteo (daily from hourly)
Open-Meteo Air Quality API (Zippenfenig, P., Open-Meteo.com, doi:10.5281/zenodo.7970649), CAMS European air quality reanalyses (validated to 2022, interim afterwards) and analyses of the CAMS regional ensemble, 0.1° resolution. · doi:10.5281/zenodo.7970649 · source page ↗
Licence : CC BY 4.0 (Open-Meteo); contains modified Copernicus Atmosphere Monitoring Service information · retrieved on Oct 25, 2026 · SHA-256 cc72a46e2da5…
Air quality of large cities worldwide, CAMS global via Open-Meteo (daily from hourly)
Open-Meteo Air Quality API (Zippenfenig, P., Open-Meteo.com, doi:10.5281/zenodo.7970649), CAMS global atmospheric composition analyses, 0.4° resolution. · doi:10.5281/zenodo.7970649 · source page ↗
Licence : CC BY 4.0 (Open-Meteo); contains modified Copernicus Atmosphere Monitoring Service information · retrieved on Oct 25, 2026 · SHA-256 027eb9b9e5fd…

© 2026 Observatoire des phénomènes naturels · About · Data sources · Manage cookies Model results and published measurements, each with its source. The Observatory’s own calculations, correspondences and forecasts, are published with their method and score.