Patel, Thomson & Wilson (2012)
Smoking increases air pollution levels in city streets: Observational and fine particulate data is a poorly conducted survey by one person, over a total period of 'over 33.6 hours' that has apparently discovered that smoking outside 'increases pollution.'
The whole abstract (from the above link)
Nowhere in the abstract or extracts do they seem to enumerate other sources of 'pollution' (car exhausts) nor do they actually seem to enumerate anything other than particulate matter as 'pollution' (carbon monoxide from any source.)
All they seem to have deduced is that 'smoking causes the product of burning leaves to locally increase where wind dispersion is so low that it can be measured.'
Another note to make is this seems to have been a single observer producing all the results. No attempt at repeatability with other observers has been made, or having multiple observers to negate any bias a single observer may introduce into the results.
Naturally, the study emphasised the 'danger' (the study failed to comment on the danger or otherwise) of particulate matter in the atmosphere as if it was there permanently and was in no way mitigated by dispersion by wind or Brownian motion.
The study is also quoted as lasting 'five-weeks' even though only a total of 'over 33.6 hours' is actually mentioned in the abstract. Since they're giving values for this number to 3 decimal places, and/or 3 significant figures, we can only assume that the most it could have been is just under 33.7 hours (or 33 hrs 42 minutes if you want it in layman's.) Clearly nowhere near the 5 weeks it's purported to take place (which is 840 hours, or to put it another way, only 4% of those five weeks was actually collecting 'evidence.')
Onto another quotation:
Significantly? Compared to 'background' traffic fumes? The study didn't cover this. Dangerous? Ditto.
That's 7 minutes 10 seconds per smoker (33.6hrs/284) presuming each smoker was a single encounter, and they weren't grouping two or more smokers together, or overlapping them. If they were doing either of the latter, then the time increases.
That number seems alarming, probably because it's the reporter who made it up by dividing some number of apples from the study, by some number of pears from a different part of the study. Or the "normal level" is such a small number that any increase would necessarily result in silly numbers when performing an "x times greater" calculation.
For example, If I have one coin and one note in my pocket and I buy something with the note, and get 9 coins in change, the number of notes in my pocket has decreased by 100%!!! Furthermore, the number of coins has increased by 9 times, or over 800%!!! Silly numbers guaranteed to make headline news.
Additionally, "peaked at" implies it didn't stay that high for very long. Little is said about how long, or how often, that peak was reached. I suspect only once throughout the whole study and it's an outlier.
Given that he wasn't measuring general pollution, merely the areas where smokers were smoking, while and shortly after they were smoking, it's hard to see how this study is forwarding any research into 'air pollution' in general.' Unless he's been measuring it over years.... oh, no - it was only 37 (I'll be generous) hours.