Originally Posted by tiptap
So I am looking at the argument between cdcox and AustinChief. And here is my take. AustinChief thinks the old Henson notion is flat out wrong because from that 1988 date we see exponential growth in CO 2 but not the corresponding increase in temperature directly correspondingly. It should have been more. He calls the more recent paper cdcox quotes as revisionist. In that it has been jiggered to be more in line with data.
So my question is does AustinChief think the generated graphs are statistical in origin or are they physical representations? A statistical representation of the earth temperatures that set the beginning temperature for every site at the average temperature of 10 degrees C or 50 degrees F. would continue over a run on a computer to continue to show that the North or South Pole was around 10 degree C. give or take the error in measurements of instrumentation usually less than .5 degrees C. However even a simple physical representation of climate systems with the same starting points at every site of 10 degrees C. would start showing the poles cooling and the equator warming. The average would still be 10 but the temperatures at the poles would greatly decrease and the equator would rise.
Is Hansen's model physical or statistical in its origin?
The actual model is physical, but the inputs and outputs are statistical in nature.
Actual future CO2 emissions were unknown. So, three statistical representations of future emissions were made: scenarios A, B, and C. Hanson argues that scenario B most closely matched the actual forcing from 1981 to 2006. Austin Chief argues that emissions exceeded those of scenario A (the case where climate forcing was projected to be the greatest). If emissions match scenario B, the Hanson model is pretty reasonable. If emissions exceed scenario A, as Austin Chief is claiming, then the predictions are pretty poor.
One last point about the statistical nature of the output. If you run a climate model from two very similar initial conditions, you will get two different projections of future climate. This is due to the chaotic nature of the physics inside the model. This chaotic behavior is also present in real world climate systems. One can run several climate simulations and determine the ensemble average and variability of future conditions. For the actual measured climate, we can only have one realization. If the actual climate measurements fall within the bounds of the ensemble average and confidence intervals of many simulations, then it can be said that the model describes the climate.
It should be pointed out that in 1988, Hansen didn't have the computing power to make ensemble runs. He could only do one run, which won't necessarily match the earth's climate in an exact manner due to the inherent internal stochasticity of the climate system.