For some time now it has been obvous that the IPCC climate models do not accurately predict the earth's temperature. This conclusion must be reached regardless of whether or not one subscribes to the catastrophic AGW hypothesis of the IPCC. Here is an analysis of the actual total heat increase of the earth compared to the predictions:

http://climatesci.org/2008/05/27/a-short-explanantion-of-why-the-monitoring-of-global-average-ocean-heat-content-is-the-appropriate-metric-to-assess-global-warming/

Can The IPCC Model Projections Of Global Warming Be Evaluated From Just Several Years Of Data?
Filed under: Climate Models — Roger Pielke Sr. @ 7:00 am
There has been an interesting and informative discussion on Prometheus regarding how the IPCC models should be evaluated. Climate Science has already shown with several examples (e.g., see) that the models are failing to skillfully predict regional climate. A test for the specific climate metric of global warming has been recommended; see

A Litmus Test For Global Warming - A Much Overdue Requirement

A Global Warming Currency

Pielke Sr., R.A., 2003: Heat storage within the Earth system. Bull. Amer. Meteor. Soc., 84, 331-335.

In the Litmus Test, I proposed the values listed below as the amount of global warming that must be achieved so as to NOT reject the IPCC claim of continuing global warming (1 * 10**22 Joules corresponds to 0.61 Watts per meter squared). This value is below the IPCC 2007 estimate of global average radiative forcing of 1.72 Watts per meter squared (see Figure SPM.2 in the IPCC SPM).

The Litmus Test

2003 8*10**22 Joules
2004 9*10**22 Joules
2005 10*10**22 Joules
2006 11*10**22 Joules
2007 12*10**22 Joules
2008 13*10**22 Joules
2009 14*10**22 Joules
2010 15*10**22 Joules`
2011 16*10**22 Joules
2012 17*10**22 Joules

These values can be compared with the best estimate of the annual average upper 700m ocean heat content change averaged over the last 4 years (which is when the data is most robust) based on the analysis of Willis et al. (2008; see) [and thanks to Josh Willis for providing this follow on analysis]. He estimates the global average warming rate for this time period, based on the upper ocean data, as -0.076 Watts per meter squared with one standard error as +/- 0.214 Watts per meter squared. This yields a best estimate of heat change of -0.48 * 10 ** 22 Joules over the last 4 years. The most positive value (using one standard deviation) is 0.88 * 10 ** 22 Joules.

Below 700 m, the heat could be accumulating, however, its pathway to reach the deeper ocean without being sampled higher up is unclear. Moreover, this heat would not be readily available to the rest of the climate system. In any case, it is hard to see how any heat change in the deeper ocean over a 4 year time period could result in large increases in the warming rate estimated in the last paragraph (ocean heat content change is an effective metric to diagnose the global average radiative imbalance as reported by Hansen et al, 2005).

Thus the value of global warming of the last 4 years fails to agree with the IPCC projections (the values are not even close!). The agrument that this is too short of a time is spurious unless the modellers can account for where else in their model results the missing Joules went.

Moreover, this is not too short of a time period to compare with the models. Heat, unlike temperature at a single level as used to construct a global average surface temperature trend, is a variable in physics that can be assessed at any time period (i.e., a snapshot) to diagnose the climate system heat content. Temperature not only has a time lag, but a single level represents an insignificant amount of mass within the climate system.

The answer to the question on this weblog “Can the IPCC model projections of global warming be evaluated from just several years of observed data” is YES. The conclusion for the past four years is that the model projections are not skillful on this time period.