Forecasting rain - Climate Change: The Facts - page 63

to GCMs that attempt to simulate and understand climate from first principles, ANNs
simply mine historical data for patterns. Many leaders within the mainstream climate
sciencecommunityaredismissiveofANNsandtheirapplication,claimingthatbecausethe
climateisonanewtrajectory,statisticalmodels,includingANNs,arenolongerapplicable.
But if through mining historical records ANNs can produce better ENSO and also rainfall
forecasts, this in itself is evidence that natural climate cycles are still operating and that the
climate is not on a radically new trajectory.
OutputfromANNsandGCMscanbeeasilyandobjectivelymeasuredusingrootmean
square error (RMSE). This number simply adds together the difference between observed
and forecast sea surface temperatures or rainfalls with the bigger the number the worse the
forecast. So it's easy to show in an objective way that ANNs can provide a much better
medium term ENSO and rainfall forecast. The difficulty has been in generating interest in
this approach and interest in the potential of ANNs to revolutionise climate science.
The future
During the 1970s and 1980s, Western democracies moved away from their traditional
methods of funding science. In Australia there was a move to project-based funding of
a limited duration. So scientists had to identify problems and promote these problems
if they were to secure funding—if they were to keep their jobs. At the same time there
was increasing interest and reliance in environmental and climate science on mathematical
modelsfacilitatinganewpreferenceforvirtualoverobservational data.In his topics cience
and Public Policy , Professor Aynsley Kellow of the University of Tasmania, explains how
this has also contributed to the hijacking of science by people grinding axes on behalf of
noble causes. 9 Richard Lindzen, Professor of Atmospheric Sciences at the Massachusetts
Institute of Technology, has repeatedly explained that climate science has now become a
source of authority rather than a mode of inquiry, that it has a global constituency and has
successfully co-opted almost all of institutional science.
Our work showing the potential application of artificial intelligence to medium-term
rainfall forecasting, recently generating five scientific publications, has been possible as
a consequence of philanthropic funding from the B. Macfie Family Foundation. The
foundation was established to provide scholars with an opportunity to seek out empirical
evidence without ideological or commercial interference. Such criteria are often critical to
the realisation of scientific research where the benefits of a line of enquiry are not always
immediately obvious and are often unpredictable. This investment needs to be leveraged
into something more significant if the big questions about natural climate cycles are to
be explored: if our initial investigations are to contribute to the development of a new
paradigm.