Heat transport to and from the mixed layer of the ocean is treated as a statistical phenomenon, enabling the development of simple regression models describing the dependence of atmospheric CO2 concentration on carbon emissions and the dependence of global average temperature on concentration. In both cases a model was found which fitted the observations according to the Ljung-Box test. From these models we were able to forecast future values of global average temperature as a function of typical “peak carbon” emission curves. The temperature curves follow the emission and concentration curves with a small delay and with maxima which are less than one degree Celsius above present day values.
This work began as an attempt to estimate Climate Sensitivity using rigorous regression methods. Climate Sensitivity is defined as the temperature response to a sustained doubling of atmospheric CO2 concentration and is used for the inter-comparison of GCMs. We soon realised that the concept itself is unrealistic because CO2 diffuses into the ocean at a rate proportional to its concentration, so
that, in order to sustain the higher concentration, a high rate of emissions would need to be sustained indefinitely. Given the finite nature of viable hydrocarbon resources, this is an unrealistic scenario.
Viewed on a time scale of centuries, human exploitation of fossil fuels in the industrial era is generating a pulse in atmospheric carbon concentration termed “Peak Carbon”. This, in turn, generates a pulse in global average temperature. The world is presently in the onset phase of this pulse; reasonable estimates of recoverable fossil fuel reserves suggest that the Peak Carbon pulse will reach a maximum within the next century or so. Global average temperature will follow suit with a maximum value which is less than 2oC above pre-industrial values.
The impulse response of atmospheric CO2 concentration is the response caused by a hypothetical, short variation in CO2 emissions. The supposed, long-lived impulse response, widely accepted by the climate modelling community, is the most egregious flaw in the application of numerical global circulation models to climate. It implies that CO2 emitted now will linger in the atmosphere for millennia. It sets the scene for the various catastrophes and tipping points presented in IPCC reports and justifies stringent emission regulations. It is based on the unwarranted assumption that GCMs provide a precise description of the ocean/atmosphere system as if it were some sort of clockwork mechanism. In contrast, our statistical model implies that the atmospheric concentration of CO2 is self-regulated by diffusion into the deep ocean and that the small perturbation of the global environment caused by the combustion of fossil fuels will be brief.
The offending paper can be downloaded here: