Updated: 80% of world CO2 emissions are related to population levels. [View all]

[GliderGuider]

Last edited Thu Oct 4, 2012, 11:54 AM - Edit history (3)

This is an OP from my comment here.

I've been wondering how much of the trouble we're in is due to population per se, and how much is due to our consumption. I've waffled back and forth on the issue. I decided to see if I could quantify the proportions of CO2 that are attributable to population and consumption.

This afternoon I developed a rough-and ready approach using normalized CO2 emission, population and "consumption" figures (more on the data later) from 1970 to 2009. Here are the steps in the analysis:

1. I normalized all data series (CO2, Population and "consumption" to be 1.0 in 1985.
   
2. In the following equation, nCO2 means normalized CO2, nPop is normalized population, and nCons is normalized consumption.
   
3. I assumed that nCO2 = xnPop + (1-x)nCons. What this means is that the proportion of CO2 associated with population growth (xnP) plus the remaining CO2 from consumption ((1-x)nCons) add up to all the CO2 we emit. I warned you it was "rough and ready"...
   
4. I calculated the expected amount of CO2 to be produced each year (1970 to 2009) while varying x from 0 to 1.0.
   
5. I found the difference between the actual and expected values of CO2, and converted that difference to a percentage of the actual value.
   
6. I graphed the percentage values.
   
7. I varied the value of x over its range until I visually found the value that minimized the excursions of the percentage variance graph.
   
8. The value of x at which the excursions were the least was deemed to be the proportion of CO2 attributable to population growth.

The CO2 data came from the BP Statistical Review 2012; the population data came from the US Census Dept.
For "consumption" proxies I used two data sets. One was global GDP in constant 1990 Geary-Khamis dollars, from Maddison. The other was the sum of the tonnes of steel and cement produced world-wide, from the USGS. I tried the latter in order to minimize the influence of the non-productive portions of GDP.

The results in both cases were broadly similar - excursions were minimized with x=0.85, meaning that over the period 1970 to 2009 about 85% of our CO2 emissions were attributable to population levels, with only 15% attributable to consumption.

The GDP plot was noisier than the steel+cement plot, as you would expect, but the overall behaviour was the same. Using the steel+cement data set, the maximum excursions were +4%/-6%. Given the level of approximation involved with this analysis and the inherent variability of human numbers and activity, that seems like acceptable accuracy.

I'm fairly confident now in saying that we would be much better served by attacking climate change through population reduction rather than consumption reduction. How that might be accomplished is, as always, a very open and controversial question.

If people want to see my work, I'll put up a web page on it.