Quote from Danilo;152241
...these numbers showed in your graph are just the legal limit values for the different Euro classes.
From what I understand these are numbers derived from model calculations but obviously related to real world values.
Real world figures seem to show for cars prior to Euro2 that the indicated numbers may even be exceeded by... a lot 
Here is the source for the above graph. Apologies for the non-scientific journal 
Here excerpts of a book with values apparently derived from measurements.
Quick summary translated (Deepl.com):
---Begin translation
Petrol engine
Let's take a petrol engine without a catalytic converter as the original state in 1985. In normal driving, it is generally estimated to emit about 1,500 to 2,000 mg NOx per kilometre. With a controlled, warm catalytic converter, on the other hand, the value usually drops to less than 100 mg, so this is the decisive step.
Unfortunately, however, this statement cannot simply be left as it is; rather, it must be supplemented by a number of important secondary aspects:
1) The older vehicles in the test (on new cars) showed many outliers towards the top. Despite the G-cat, four-digit NOx values were occasionally measured. The probable cause was thought to be sloppily programmed regulations outside the test cycle. Since EURO 1, this has often affected speeds above 120 km/h. In the case of even older cars with a US catalytic converter, on the other hand, the test cycle ended at 96.4 km/h, which is where a particularly large number of anomalies occurred. Over the years, the proportion and the amount of these outliers decreased considerably.
2) The 100 mg were not the end of what was technically possible. With each newer standard, emissions fell further and further in operation with functioning regulation. Cautiously moving current models (EURO 5) tend to fall into the single digits or right to the detection limit. In relative terms, therefore, emissions differ greatly within petrol engines with a warm G catalytic converter, but in absolute terms it is quite irrelevant whether one emits 10 or 100 mg. The crucial thing is that the 2,000 mg is off the table.
3) The petrol engines before EURO 3 (the ones with the 40-second lead time in the measurement) still have a hard time after a cold start. As a rule of thumb, you have to reckon with about a full kilometre until the catalytic converter has kicked in. Since EURO 3, on the other hand, it usually takes just a few seconds for the exhaust gas purification system to kick in.
Since the introduction of EURO 3, NOx emissions from petrol engines in real mixed traffic have been 100 mg or less. This should be a safe level for air quality.
In the case of our old cars, on the other hand, which are typically labelled EURO 1 or EURO 2, it is not easy to give a blanket figure for nitrogen oxide emissions, and in fact it does not make sense to do so. Depending on the length of the journey and the speed, a mixed calculation of warm phases (< 100 mg) and cold phases (> 1,000 mg) must be made, taking special account of any inadequate regulation. As an average value in mixed traffic, 500 mg for EURO 1 and 300 mg for EURO 2 can be assumed.
----- End translation
So.... if BMW jots down a big "O" it was probably meant to be read as "O"oooh we really don´t know 
Reinhard
) fand das Eigelb irgendwann einmal zk und Teil-Toledoblau viel schöner.
