Normalizing CO2 emissions makes little physical sense (at least to me)

I have written about some of the problems of normalizing CO2 emissions (here, here, and here), and I don't really understand the reasoning behind it, other than to make a policy statement. Indeed, because we live in a world that has dominant social systems that govern the physical systems, we must recognize the importance that such framing has in policy statements and decision-making.

We must recognize that China - although the world's largest emitter of CO2 - has a far lower per capita CO2 emission than the United States. However, the physical world doesn't care about this normalization, and as China has pursued their economic growth through completing about 1 coal fired power plant each week, its economic costs due to coal and industrial pollution are draining an increasing amount away from that growth.

We must recognize that countries that divide their CO2 by their GDP also show some sort of "CO2 efficiency" in their economy. However, the physical world doesn't care about this normalization or any concept related to it.

To the physical world, one ton of CO2 emissions is the same as any other single ton of CO2 emissions, whether it is emitted from China, the US, or Botswana; whether it's emitted due to respiration, cooking fires, or industrial emissions; whether it's for economic gain in country A or in country B. An economist might look at this and say, "Hey, this is a great reason for applying an opportunity cost assessment here," but - you know what? - the physical world doesn't care about opportunity costs, either.

With all that recognition, we now have a new study - from the University of Michigan's Transport Research Institute - that shows that the US's CO2 emissions don't fare "as bad" when you account for heating degree days (and cooling degree days).

Sivak and UMTRI colleague Brandon Schoettle say that some rankings adjust the total amount of emissions to account for the size of each country's population (per capita) and its overall economic output (per GDP). But they believe it's important to go one step further: account for the general heating and cooling demands imposed by the climate of a given country because climate control produces carbon dioxide emissions.

The justification for this is not really clear. Are they doing this based on the assumptions that we will have climate change, and then people might be using energy differently? If so, the article doesn't actually say how or what climate change scenarios are assessed. If it's looking at how current energy is used based on current climate, then it's double-counting types of efficiency, since part of those costs would be included in GDP measures.

What gets me most, though, is the final paragraph of the article:

"Overall, our results suggest that taking climate into account makes a significant difference in how countries fare in carbon dioxide emissions rankings," Sivak said. "Because people respond to the climate they live in by heating and cooling indoor spaces, an index that incorporates climate provides a fairer yardstick than an index that does not."

Ummm.... How does Sivak define, "fairer?" If one is concerned about total emissions' impacts on the world and on global climate change, then his argumentation makes no sense, because the physical world doesn't care about this normalization. In fact, as I stated above, the whole idea of normalizing based on per capita or per GDP makes little physical process sense, except for the arguable forum of of public policy. However, that forum must divorce policy making and future scenarios from impacts from climate change in order for the argumentation to really work outside of an international setting. However, recognizing that policy may be divorced from the impacts of climate change doesn't justify normalizing by degree days.

Why not, for example, choose something else, like normalizing by weather variability (which is connected with major changes in local climate), by expected health care costs (since this has major direct and indirect impacts on GDP and per capita as well as speaks to a host of environmental changes), by security (since we can expect major changes in migration - and in nations' security - in a future with climate change), or by almost anything else than a triple normalization that included degree days?

Or maybe I'm missing some justification that makes sense that was presented in the article in Scientific American that I didn't see in the PhysOrg article. The way that it was presented in PhysOrg made it seem more like a "hey, don't hate on the US, we're better than you think when it comes to CO2 emissions*".

* When you normalize the data three ways to next century.