Skip to content

Instantly share code, notes, and snippets.

What would you like to do?
NO2 in Munich, 2016

NO2 in Munich 2016: high traffic => high NO2


This plot shows NO2 levels over the day in Munich in June and December 2016. München-Landshuter-Allee on the left has about the highest NO2 levels in all Germany, and a lot of traffic — 120,000 to 150,000 cars and light trucks per day.
Surprise: high traffic => high NO2.

(What are the little dots in the plots ? Each dot is NO2 at a given hour and day in June or December (actually, a 1-hour average). Each vertical line, 0:00 to 23:00, has 30 red dots for NO2 in June, and 31 blue dots for NO2 in December. The red curve is the average of the red June dots, the blue curve of the blue December dots. The 5 short lines on the left of each panel mark the 10, 25, 50, 75, 90 percentiles, a.k.a. a rug plot: 10 % of the 24 x (30 + 31) measurements are below the bottom line, 10 % above the top line.)

Dieselverbot: run a 30-day trial ?

Some believe that prohibiting non- Euro6 diesels in inner cities would reduce NO2. How could one test and quantify this belief: 5 % less NO2, 10 %, can't say ? Any engineer would run a trial:

  • allow only Euro-6 diesels in a few inner cities for say 30 days
  • measure NO2 etc.
  • compare the values with previous months / previous years.

Does anyone know if such trials have been run, anywhere in Europe ? With open data on the web ?

A clear decision process for the results of the trial is essential. Just for example,
> 10 % improvement: more Dieselverbot
< 10 %: not worth it, rethink.
Given the emotional non-debate on Dieselverbot in Germany, this seems unlikely.

How much less NO2 ?

A question for statisticians: These time series of NO2 vary from night to day, winter to summer, and across the city. Suppose a trial Dieselverbot for 30 days gives us new time series, 24 x 30 x 5 numbers, and the claim "p % less NO2" than in previous months / years. How would one go about evaluating such a claim — what model, statistics methodology, computer programs ? Even defining "p % less" is daunting: what does the average person breath, over the day ?

Trials of this kind must be common in medicine and public health; I'd welcome links to tutorials.

Why NO2 limits 40 ug/m3 in the EU, 100 in the US ?

A question for toxicologists: the EU standard limit for NO2 is a yearly average of 40 ug/m3; in the US, the yearly EPA standard is 53 ppb ~= 100 ug/m3. Why are these standard limits so different ? I realize that setting limits is not easy, but still.


A note to newspapermen, doomsayers, scandalmongers: reporting only the highest values in the whole country, Landshuter Allee or Neckartor, is imho one-sided and misleading.

Comments are welcome, more data is welcome

I'd be particularly interested in NO2 and ozone data at airports, train stations, and highways; also in data at measurement stations near each other — how does NO2 spread out and dissipate ?

-- denis

Last change: 2017-08-24 August

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment