Are there basic physical constraints on future anthropogenic emissions of carbon dioxide?

Global Climate Models (GCMs) provide forecasts of future climate warming using a wide variety of highly sophisticated anthropogenic CO2 emissions models as input, each based on the evolution of four emissions "drivers": population p, standard of living g, energy productivity (or efficiency) f and energy carbonization c. The range of scenarios considered is extremely broad, however, and this is a primary source of forecast uncertainty. Here, it is shown both theoretically and observationally how the evolution of the human system can be considered from a surprisingly simple thermodynamic perspective in which it is unnecessary to explicitly model two of the emissions drivers: population and standard of living. Specifically, the human system grows through a self-perpetuating feedback loop in which the consumption rate of primary energy resources stays tied to the historical accumulation of global economic production - or p times g - through a time-independent factor of 9.7 +/- 0.3 milliwatts per inflation-adjusted 1990 US dollar. This important constraint, and the fact that f and c have historically varied rather slowly, points towards substantially narrowed visions of future emissions scenarios for implementation in GCMs.