Following Frank's post on Dwight Lee's fine AJC column (whenever I read a piece by Dwight, I'm impressed by how easy he makes it seem to say important things well), I'd like to add a couple of comments from Peter Huber. Huber points out, here, that another way (in addition to reducing flatulence) that internal combustion engines have helped is by letting farmland return to the forests:

Happily, however, our agricultural footprint has been shrinking a lot faster than our cities have been sprawling. When Europeans first arrived on this continent, the area now represented by the lower 48 United States had about 950 million acres of forest. That area shrank steadily until about 1920, to a low of 600 million acres, as Americans spread across the landscape.

Then, astonishingly, we began to retreat, and the wilderness began to expand once again. Precisely how fast is hard to nail down: The continent is large, most of the land is privately owned, and the definitional debates rage. But all analyses show more, not less, forest land in America today--somewhere between 20 million and 140 million acres more--than in 1920. Roughly 80 million more acres of cropland were harvested 60 years ago than are harvested today.

Indeed, an implication of this change is that North America may be a net carbon sink, unlike some signatories of the Kyoto Accord. Huber again:

As best these things can be measured, the North American carbon books are now in balance. Carbon dioxide levels downwind of the estate, out in the Atlantic, are lower than upwind, in the Pacific. Carbon-sink skeptics say they don't see enough new trees to account for the drop. But then, global warming skeptics say they don't see enough human carbon emissions to account for rising temperatures. The weight of the evidence indicates both a warming planet and a huge North American carbon sink. The carbon-sink data are, if anything, the more reliable, because they require only direct measurement today, not estimates of conditions a century ago.

If we can't precisely explain where all the carbon is sinking, it's because it's hard to track deposits that average 0.002 inch over a vast continent. Many forest inventories count only "lumber quality" trunks, ignoring younger trees and grassland. New forests mean new, carbon-rich soil, which is almost impossible to inventory accurately. New soil means new sources of silt in rivers, which dump carbon into the ocean. Pampers sequester carbon, too, as long as the landfill they wind up in doesn't permit them to decompose.