~ $ cd projects/no2 && cat README.md

NO2 by the hour and the month

Nitrogen dioxide in a city is mostly traffic. The Generalitat de Catalunya measures it every hour at fixed points and publishes every reading since 1991. Looked at a day at a time it is noise. Averaged by hour of the day and month of the year, it is a picture:

Barcelona (Poblenou) · background, urban · mean NO2 in µg/m³ by hour of the day and month of the year · every day of the week, 1991–2025
JanFebMarAprMayJunJulAugSepOctNovDec
01485555443937313043524949
02435052444238333243484544
03404549424237343340444041
04364144414038353338393738
05343842383938353136363535
06354042404040393337373636
07404447464646463842414040
08475254535253534449484846
09535758575657574653525350
10525657565353534453535451
11515552504644433749535250
12525243413734312941495252
13504636343129262332424951
14473931292625221825364649
15433528242321181521314146
16423528242321191520304245
17463831242421201522324648
18544635262723211624375657
19615541292926231827426362
20636047333329262130486664
21636453393732282435526665
22616556423934302637556463
23586357434034312738566059
24535957433935302841555453

The mean of each year, every day of the week

0102030405060708019952000200520102015202020251991: 43.5 µg/m³1992: 40.3 µg/m³1993: 35.7 µg/m³, from only 69% of the year's hours1994: 46.2 µg/m³, from only 59% of the year's hours1995: 48.5 µg/m³, from only 71% of the year's hours1996: 45.0 µg/m³1997: 50.7 µg/m³1998: 59.2 µg/m³, from only 74% of the year's hours1999: 73.0 µg/m³, from only 25% of the year's hours2000: 59.8 µg/m³, from only 52% of the year's hours2001: 42.0 µg/m³2002: 53.3 µg/m³2003: 57.4 µg/m³2004: 40.3 µg/m³2005: 43.5 µg/m³2006: 47.6 µg/m³, from only 38% of the year's hours2007: 42.4 µg/m³, from only 72% of the year's hours2008: 47.6 µg/m³2009: 50.8 µg/m³2010: 45.6 µg/m³2011: 43.1 µg/m³2012: 42.3 µg/m³2013: 40.0 µg/m³2014: 39.3 µg/m³2015: 45.2 µg/m³2016: 42.8 µg/m³2017: 44.1 µg/m³2018: 38.7 µg/m³2019: 36.7 µg/m³2020: 29.0 µg/m³2021: 25.7 µg/m³2022: 27.2 µg/m³2023: 24.3 µg/m³2024: 23.7 µg/m³2025: 22.0 µg/m³EU limit, 40WHO guideline, 10

Source: Generalitat de Catalunya, Xarxa de Vigilància i Previsió de la Contaminació Atmosfèrica. Dades obertes. The dataset, at its source. This site keeps sums of the finished years 1991 to 2025, last added to on 21 September 2026.

Each cell is the mean, in µg/m³, of every measurement taken at that hour in that month over the years selected. The colour says whether it is good or bad before the number is read: green is clean, red is the European annual limit of 40, and purple going to black is beyond it. Press a bar to see one year alone.

What is in it

At Barcelona's Poblenou, a station away from the big roads, over its whole record:

Choose the Eixample, which measures traffic, and the same shape is there in purple: 70 µg/m³ in 1999, 49.5 in 2019, 35.2 in 2020. The year of the lockdowns was its first under the limit; 2022 went back over it, and 2025 closed at 29.1. The same shape, higher or lower, is at every urban station here.

Height

Choose the Observatori Fabra. It is in Barcelona, 415 m up the hill the city climbs, and its table is green: 7.4 µg/m³ in 2025, when the Eixample below it closed at 29.1 and Poblenou at 22.0. It has no rush hours either. Its highest cells are late morning in summer and late afternoon in winter, and none of them reaches 20; its lowest are the early morning, just when the streets below are filling. Same city, same traffic: little of it is measured up there.

The other green table is La Castanya, in the Montseny, where the yearly mean has stayed under 4 µg/m³ since 2013: that is what air with no road under it measures. Those two are the only stations of these eleven that ended 2025 under the 10 µg/m³ the WHO has recommended since 2021.

What it is not

The cells are means, and a mean hides the bad days. Public holidays are counted as whatever day of the week they fell on. The dataset numbers its hours 01 to 24 and does not say by which clock. A year drawn as an outline had less than three quarters of its hours measured; Eixample has no 2010 at all. These are open data added up, not validated science: for anything that matters, go to the source.

Where it comes from

I first drew this table in 2021, for myself, in a small app that asked the portal for everything on every visit. This page is that analysis again, with the asking moved out of the reader's way: the site keeps the sums, a finished year at a time, and asks the portal only when a year has ended.

~/projects/no2 $

~/projects/no2 $