Conclusions - Climate Change: Identification and Projections - page 132

because they are static - the climatic cause-effect relationships are dynamic
by their very nature. The theory of identification helps to directly handle this
situation. It is not only effective, but indispensable, and also allows us to
deal with the multi-causal nature of the climatic process other than with
multiple static regressions. We will not reiterate the theoretical aspects of
identification here, but as a reminder of the results: the analysis of
uncertainty and hypothesis tests show that the contribution of solar activity
to the evolution of global temperature is not a simple effect resulting from
climatic variability. Instead, its level is, with high probability, up to ten times
greater than the insignificant level assigned to it by the IPCC, on the basis of
a single energy factor 3 . This probability is not just an “expert opinion”, but is
based on actual statistical calculations. It must be considered as a fact that
solar activity, as a causal explanatory variable, does constitute the primary
deterministic explanation of “climate change”. Understanding the reasons for
this action (cosmic activity or other) is another issue and does not fall within
our remit.
12.4. Predictive capacity
In all case studies examined, the analysis of the simulations' output error
shows that random climate fluctuations greatly contributed to the warming
over the last half-century, whereas the IPCC would like to use natural
variability to explain the current stagnation. The predictive power of the
identified models confirms that the first hypothesis is correct one. Beyond
accuracy calculations, nothing rivals the capacity of a model at predicting
future behavior. Traditionally, for identification of industrial processes, a set
of input-output data is kept for validation purposes. In this case, the
experiment consists of carrying out identification on the database recorded
until 2000 only, then comparing real temperatures which followed on with
simulations. The test is conclusive: despite the uncertainty for identified
parameters, the model reproduces the current temperature stagnation since
the year 2000, even though nothing seemed to indicate such an outcome. In
this regard, not one of the IPCC models made such a prediction, neither of
the stagnation itself, nor of anything resembling it. In the possible scenario
3 Unless the solar irradiance reconstructions accepted by the IPCC are substantially reviewed,
as Shapiro proposes.