Answer by Francis Chen:
As touched upon already, air pollution is a consequence of massive resource consumption, and there are different sources of air pollution, not just cars.
However, I'm going to speak from a North American perspective (Hong Kong, aspoints out, has different sources of air pollution, given 90% of people take transit there already).
I argue that land use, not public transit, plays a bigger factor in reducing greenhouse gas emissions. Without proper land use, you can have lots of buses go any where, but they wouldn't be able to meet the right ridership since everything is far apart.
Aspoints out, land use is very critical. Your goal: you want to be able to reduce the travel time from your origin to your place of destination, so you choose the mode of transportation that is the least costly, from a time and monetary perspective. You want to reduce the time it takes to take to get to work or buy groceries.
For example, if you can make the places between where you live and the places you want to go (grocery store, work, park, etc.) closer to each other, you make it a lot easier to depend on the car, since things are not spread out. If things are too spread out, the Golden Triangle of Transit here fails:
From Jarrett Walker, international transit consultant: ().
In short, without the right density and walk-ability, transit services can't justify frequent services (buses coming every five minutes!) or have the right ridership to meet the service. This is why in a lot of suburbs, you have buses coming every 30 minutes or hour: it is much more cheaper for them to take that route because not many people take transit.
Without proper land use planning (which is a whole other subject on its own, as the comments on Seth's post point out), like in the US, you have a lot of suburbs, which makes it more convenient to drive than to take transit.
Jarrett Walker, from his book, Human Transit: How clearer thinking about public transit can enrich our communities and our lives (see here:) notes that if land use ("development") is spread out, transit would not be very efficient:
The red line is the transit line, the dark purples are the first and last stops, and the light purple are the intermediate stops.
If we pretended, this was a suburb, having cul-de sacs (in light purple) would not be very efficient for transit vehicles because the bus/rail would have to go through every destination. It would take a really long time for that red line to go to every place.
This is a different transit line, but much more straight forward.
Now, with the proper land use that is very dense in one line, this can be a very efficient transit line. This is key: proper land use. You can still have a straight line but it can be located in the suburbs, and it would be an inefficient transit line because there is not enough ridership
Here's an example of density, visualized, from the architecture firm, Perkins & Will, of a urban design plan for Vancouver, Canada, I believe.
Here's another similar image of Vancouver, probably at a different location:
Watch that transit line pass through that dense neighborhood! With the amount of development, it will have a justifiable amount of ridership to justify frequent service along the line to take transit!
So, in short, land use, not public transit, in North America, has a greater factor in influencing greenhouse gas emissions.
Note: you might have noticed that I was sort of repetitive in different ways. It is really important to be clear when discussing or talking about transportation and the language of transportation because people interpret things different ways. That's why I provided lots of different visuals from different sources to help put everything in context.
The concept of density, for example, is a very broad concept. Does it mean 5 story, 10 story, 80 story (Manhattan), etc. There are different ways of measuring high density. Similarly, the concept of mixed land use or closer land uses is also very hard to visualize and thus has to be explained in depth to the reader.
This is the role of urban planners, transportation engineers, and also to a growing extent, the smart cities / big data movement, which I talk more about in this article: