What to make of quotes by smart folks

I was reading an article on Vernier's (a maker of physics and chemistry laboratory instruments and software primarily for high school physics students) web site about an experiment in which their Labquest unit and some of their temperature probes were used in an project to determine the effect of roof color on the temperature of a dwelling.

In the article, I read that Energy Secretary and Nobel Prize laureate, Dr. Steven Chu had stated that "Giving roads and roofs a paler color would have the same effect of removing every car in the world off the roads for 11 years." My eyebrows went up for a couple of reasons. First, I have a hard time understanding what is even being claimed. Second, any interpretation I was able to make seemed implausible.


I used that quote as a Google search phrase and found quite a few places repeating the quote but the best I could do for an original source was in the Telegraph, a British newspaper. There, the amplified quotation was "If you look at all the buildings and if you make the roofs white and if you make the pavement more of a concrete type of colour rather than a black type of colour and if you do that uniformally (sic), that would be the equivalent of... reducing the carbon emissions due to all the cars in the world by 11 years – just taking them off the road for 11 years,"


I'm sorry, but that is just a very confusing statement. Is the implication that making this change immediately and for all time would cumulatively equal the amount of carbon emissions due to the world's cars for 11 years? The confusion stems from having a time frame for the cars but not for the houses and buildings. This is, in my opinion, more unit ambiguity, and Dr. Chu has a Nobel Prize! Further, is that all the cars on the road today, or the cumulative emissions of those cars plus the cars to be added over the next 11 years?


I suspect that rough numbers will suffice to see what Dr. Chu might be talking about and whether it makes any sense. So just how much carbon is emitted by all the world's cars today? Here I find the following:

"The authors present the image of a coal train that stretches 55,000 miles, long enough to circle the globe twice, carrying 314 million metric tons (314*10^6 tonnes) of carbon - the amount of CO2 emitted by U.S. cars and trucks in the year 2004."

Argh. "314 million metric tons of carbon" and the amount of "CO2 emitted." CO2 is 72.7% oxygen and 27.3% carbon by weight. So are they referring to 314/.273 or 1150 million tonnes of CO2 containing 314 million tonnes of carbon or is it 314 million tonnes of CO2 containing 85.7 million tonnes of carbon? Sneaking in the back door by looking at annual gasoline consumption, assuming gasoline is more or less n-heptane which is 84% carbon by weight approximated the carbon to 316 million metric tons and that's closer than I deserve so I'm going with 314*10^6 tonnes of carbon from 1150*10^6 tonnes of CO2. The article cited states that the U.S. contributes 45% of the world's automotive CO2 emissions so the world's total is estimated at 2555*10^6 tonnes of CO2. This is for 2004, so we multiply by 11 to see what Dr. Chu was talking about. This yields 28.10*10^9 tonnes of carbon.


Now, what about roofs and pavements? This is a dramatically more difficult estimate and I'm going to rely on the scholars. Here is the best article I could find without paying. An abstract of an original paper (the earlier of the two cited in the article above) is at Springerlink and for $34 you can have the whole paper. I'll quote the abstract here:

Increasing urban albedo can reduce summertime temperatures, resulting in better air quality and savings from reduced air-conditioning costs. In addition, increasing urban albedo can result in less absorption of incoming solar radiation by the surface-troposphere system, countering to some extent the global scale effects of increasing greenhouse gas concentrations. Pavements and roofs typically constitute over 60% of urban surfaces (roof 20–25%, pavements about 40%). Using reflective materials, both roof and pavement albedos can be increased by about 0.25 and 0.15, respectively, resulting in a net albedo increase for urban areas of about 0.1. On a global basis, we estimate that increasing the world-wide albedos of urban roofs and paved surfaces will induce a negative radiative forcing on the earth equivalent to offsetting about 44 Gt of CO2 emissions. At ∼25/tonne of CO2, a 44 Gt CO2 emission offset from changing the albedo of roofs and paved surfaces is worth about 1,100 billion. Furthermore, many studies have demonstrated reductions of more than 20% in cooling costs for buildings whose rooftop albedo has been increased from 10–20% to about 60% (in the US, potential savings exceed $1 billion per year). Our estimated CO2 offsets from albedo modifications are dependent on assumptions used in this study, but nevertheless demonstrate remarkable global cooling potentials that may be obtained from cooler roofs and pavements.

Note that the abstract doesn't mention a time period over which the 44 Gt (gigatonne, or billion metric tons) offset occurs. The article linked above seems to clarify it as the total offset, and splits the offsets into 24 Gt for roofs and 20 Gt for pavements, along with the assumptions as to the increases in albedo necessary. And the abstract is very clear that they are discussing tonnes of CO2, not carbon, so we have to to take 27.3% of the 44 Gt for a total reduction of carbon of 12 Gt. This is less than half of the 28.1 Gt I found for worldwide automotive carbon emissions for 11 years. Still, it's a surprisingly large number, much more than I'd have estimated off the top of my head (like Mr. Spock Sherlock Holmes, I never guess).