Atlanta and Barcelona are roughly equal in population, but they are not equal in size. The two metropolitan areas, which each contain around 5 million people, look about like this from a bird's eye view:

And, as you can see in the graphic from the World Resources Institute, carbon emissions from public and private transportation combined are dramatically different, counted in tons of carbon dioxide per person. These two characteristics — the literal footprint of a city and the carbon footprint of its transportation — are intimately linked. The more spread-out an urban area, the more likely its residents are to run even the most routine errands by car, producing vehicle emissions. The more compact it is, the less distance residents need to travel every day, and the easier — and cheaper — it is to build public transit.

These two cities illustrate a big piece of the climate change picture that gets much less attention than coal plants or hybrid cars: As the world urbanizes at break-neck speed, the way we design growing cities will heavily determine the environmental impact of the people who live there. And decisions we make today about where and how to invest in transportation will lock in those impacts for decades.

If the world's cities develop like sprawling Atlanta, the area of urbanized land on the planet could triple from 2000 to 2030, according to the World Resources Institute report. The number of cars could double to 2 billion, as could the amount of land used globally per household. The problems that would then arise would be both environmental and economic. We'd have to spend a lot more money paving roads and extending utilities to people who live farther apart than in close quarters. Houston, WRI points to as an example, spends about 14 percent of its GDP on transportation. Compact, bike-happy Copenhagen devotes about 4 percent.

The alternative to Atlanta doesn't have to look like Manhattan (Barcelona is nowhere near as dense as that). And it would be hypocritical of Americans who already enjoy extensive road infrastructure, ubiquitous cars and spacious homes to now lecture the developing world that it can't have the same. The fundamental issue here, though, isn't about forcing everyone into high-rises or out of their cars and onto bicycles. It's about planning for the growth to come instead of simply letting it happen haphazardly.

Invariably, the latter leads to sprawl, congestion, wasted productivity and higher infrastructure costs. But if we plan now to accommodate millions more global city-dwellers in more compact cities, with better transit infrastructure, we can rein in the massive costs — for the environment and local economies — associated with unchecked, infinite urban development. Do this in the world's 724 largest cities, WRI says, and we might curb greenhouse gas emissions by 1.5 billion tons a year by 2030. If we built denser, more connected cities, the report adds, we could save $15 trillion in infrastructure investments over the next 15 years.

To put this in further perspective, the Institute for Transportation and Development Policy and UC Davis recently modeled what might happen if global cities heavily shifted away from constructing new roads and parking garages and subsidizing gas consumption, and focused instead on expanding public transit and non-motorized transportation (inherent in this shift is the idea that we'd have to build more compactly). Do that on a massive scale, Michael Replogle and Lewis Fulton concluded, and we could save $100 trillion in public and private capital and transportation operating costs by 2050. We could reduce emissions from urban passenger transport by 40 percent.

Change the ways we get around, and how much we travel, and an alternative future come 2050 would look very different than if we continue with business as usual:

Translate that into emissions, and we get this: