We've come to expect that the "solutions" to Climate Change will be high tech--floating windmills, underwater generators, and nano solar in the Jules Verne/James Bond tradition--or at least high concept (carbon credit trading). But this week Energy Secretary and Nobel Prize winner Steve Chu is in Europe extolling the benefits of ... white roofs. The concept is simple but the numbers he cites are massive: Making roofs and pavement more reflective could offset 44 billion tons of CO2, or the equivalent of taking all of the world's cars off the road for 11 years. (While these numbers appear huge, there's no mention of the time frame, so they're not comparable to other numbers.)

Even so, these numbers are an interesting reflection of how much we've re-engineered the planet and climate already, and how we might start to mitigate that. Chu seems to want us to understand this as a category-jumper, working on climate in a different way than changing energy sources or sequestering carbon, and so he describes increasing the albedo (reflectivity) of buildings and pavements as "geo-engineering."

I guess it's predictable that some media are already making fun of Chu. Perhaps the idea is too simple and sensible, or just too eggheaded. (In California a related proposal to require that car roofs be more reflective as part of the state's climate policy became controversial. There's something about the brainy dumbness of the idea that doesn't float politically.) Over at google's geo-engineering forum, though, they're worried that the reception for Chu's idea is an indication of the ridicule the public may heap upon heavier geo-engineering against climate change--things like stratospheric aerosols and cloud whitening. A few years ago, discussing geo-engineering in polite company was pretty much off limits because it muddied up the discussion about when to take action on climate change. But as action has been delayed, and fear has grown, the concept has moved steadily towards the far outskirts of the center. Chu's use of the term geo-engineering in this relatively benign context is an indication of where the discussion is headed.