Unmapped Dataworks

Free tools · From a European air-quality model

Air in Mafube

Air in 2025 (modelled): nitrogen dioxide near the ground averaged 20 µg/m³, 2.0 times WHO’s guideline of 10; winters 26 µg/m³, summers 15 µg/m³.

Nitrogen dioxide (NO2) comes mostly from burning: power stations, vehicles, industry. Near the ground is what people breathe. The column is all of it in the air above, from the ground to the top of the sky, which is what a satellite sees. Sulphur dioxide (SO2) comes mostly from burning coal. Ozone forms in sunlight from other pollutants.

Modelled, not measured: these figures come from the Copernicus Atmosphere Monitoring Service’s reanalysis, a weather and chemistry model that takes in satellite readings. Its cells are about 80 km across, so a busy road or a smoky winter evening in one suburb is smoothed into the region around it. It is not a reading at any address.

2025

WHO’s 2021 guideline for nitrogen dioxide is 10 µg/m³ as a yearly mean. It is a long-term health guideline, not a legal limit (South Africa’s national standard for the yearly mean is set separately), and these are modelled regional means, not readings where you live.

Month by month

MonthNO2 near the groundOzone near the groundNO2 columnSO2 column
December 202516 µg/m³57 µg/m³94 µmol/m²111 µmol/m²
November 202514 µg/m³57 µg/m³104 µmol/m²106 µmol/m²
October 202519 µg/m³61 µg/m³121 µmol/m²130 µmol/m²
September 202519 µg/m³59 µg/m³126 µmol/m²155 µmol/m²
August 202528 µg/m³51 µg/m³148 µmol/m²167 µmol/m²
July 202529 µg/m³41 µg/m³155 µmol/m²171 µmol/m²
June 202525 µg/m³40 µg/m³154 µmol/m²146 µmol/m²
May 202524 µg/m³39 µg/m³150 µmol/m²144 µmol/m²
April 202523 µg/m³37 µg/m³123 µmol/m²107 µmol/m²
March 202519 µg/m³51 µg/m³109 µmol/m²114 µmol/m²
February 202513 µg/m³51 µg/m³104 µmol/m²120 µmol/m²
January 202515 µg/m³53 µg/m³96 µmol/m²102 µmol/m²
December 202414 µg/m³61 µg/m³102 µmol/m²125 µmol/m²
November 202418 µg/m³66 µg/m³119 µmol/m²136 µmol/m²
October 202416 µg/m³71 µg/m³115 µmol/m²151 µmol/m²
September 202418 µg/m³64 µg/m³117 µmol/m²144 µmol/m²
August 202422 µg/m³54 µg/m³125 µmol/m²138 µmol/m²
July 202425 µg/m³41 µg/m³131 µmol/m²108 µmol/m²
June 202428 µg/m³38 µg/m³166 µmol/m²160 µmol/m²
May 202430 µg/m³43 µg/m³145 µmol/m²167 µmol/m²
April 202420 µg/m³42 µg/m³127 µmol/m²140 µmol/m²
March 202420 µg/m³50 µg/m³132 µmol/m²163 µmol/m²
February 202417 µg/m³52 µg/m³103 µmol/m²134 µmol/m²
January 202415 µg/m³58 µg/m³78 µmol/m²98 µmol/m²

Over the last five years the NO2 column averaged 153 µmol/m² in winter against 99 µmol/m² in summer, and the SO2 column 156 µmol/m² against 115 µmol/m²: still winter air holds what is burnt close to the ground, and more is burnt to keep warm.

Every year since 2003

2003: 14 µg/m³2004: 14 µg/m³2005: 14 µg/m³2006: 14 µg/m³2007: 15 µg/m³2008: 16 µg/m³2009: 16 µg/m³2010: 17 µg/m³2011: 16 µg/m³2012: 16 µg/m³2013: 16 µg/m³2014: 18 µg/m³2015: 18 µg/m³2016: 18 µg/m³2017: 18 µg/m³2018: 19 µg/m³2019: 20 µg/m³2020: 19 µg/m³2021: 19 µg/m³2022: 20 µg/m³2023: 20 µg/m³2024: 20 µg/m³2025: 20 µg/m³200320082013201820232025
Nitrogen dioxide near the ground, each year’s mean (modelled). The line is WHO’s yearly guideline, 10 µg/m³.
Show the numbers
Nitrogen dioxide near the ground, the mean of each year, 2003 to 2025: every figure
When or whatFigure
200314 µg/m³
200414 µg/m³
200514 µg/m³
200614 µg/m³
200715 µg/m³
200816 µg/m³
200916 µg/m³
201017 µg/m³
201116 µg/m³
201216 µg/m³
201316 µg/m³
201418 µg/m³
201518 µg/m³
201618 µg/m³
201718 µg/m³
201819 µg/m³
201920 µg/m³
202019 µg/m³
202119 µg/m³
202220 µg/m³
202320 µg/m³
202420 µg/m³
202520 µg/m³

See it in the Copernicus Browser, the European Space Agency’s free viewer, on the newest image: Sentinel-5P, nitrogen dioxide · Sentinel-5P, sulphur dioxide. These are Sentinel-5P’s observed columns for one day, not our modelled yearly figures; days with cloud have gaps.

Every municipality, ranked · Everything we hold for Mafube · On the explore map

Source

Modelled, not measured: these figures come from the Copernicus Atmosphere Monitoring Service’s reanalysis, a weather and chemistry model that takes in satellite readings. Its cells are about 80 km across, so a busy road or a smoky winter evening in one suburb is smoothed into the region around it. It is not a reading at any address.

How we worked this out

What we changed: we averaged the reanalysis’s monthly means over each place by its area in each 0.75 degree cell, turned NO2 and ozone on the model’s lowest level from mass mixing ratio into micrograms per m3 with the month’s air density (surface pressure and temperature), and the columns from kg per m2 into micromol per m2.

Particles (PM2.5, PM10) and sulphur dioxide near the ground are not shown yet: the model reads them far higher over Gauteng and the Highveld than is plausible for a regional mean, so they wait for a check against ground stations.

Contains modified Copernicus Atmosphere Monitoring Service information 2003 to 2025. Neither the European Commission nor ECMWF is responsible for any use that may be made of the Copernicus information or data it contains. Copernicus Atmosphere Monitoring Service (2020): CAMS global reanalysis (EAC4) monthly averaged fields. Copernicus Atmosphere Monitoring Service (CAMS) Atmosphere Data Store, DOI 10.24381/fd75fff2. Inness, A. and others (2019): The CAMS reanalysis of atmospheric composition, Atmos. Chem. Phys., 19, 3515 to 3556. Under CC BY 4.0.

Dataset: air quality by municipality, release of 12 November 2026.

Unmapped Dataworks