Forecasting global climate change - Climate Change: The Facts - page 133

Our forecasts for each year's global average temperature for the 100 years to 2113 are
thattheywillbethesame,moreorless,asthe2013globalaveragetemperature.Wesuggest
that our forecasts should be monitored against the University of Alabama at Huntsville's
(UAH) lower troposphere temperature series because this satellite-based measure provides
abetterassessmentoftheglobalaveragethantheHadley(HadCRUT3)series,andbecause
it is fully and openly documented and is, therefore, less likely to be biased.
Perhaps it is possible to improve on the already very accurate longterm temperature
forecasts from the no-trend model, for example by estimating the global average
temperature level from a weighted average of temperatures over recent years, rather
than from only the latest year. We have not attempted to improve upon our very simple
no-change model, however, because the errors of the forecast from the model are too
small to be of concern to policy makers and business decision makers; the no-trend model
forecasts are more than good enough.
What do previous environmental alarms tell us?
Having investigated the forecasting procedures behind the IPCC forecasts and found them
to lack validity, and having found the forecasts to be much less accurate than no-change
forecasts, we were concerned that governments are taking the dangerous man-made global
warming alarm seriously, to the extent that they have already implemented costly policies
andregulations.Wedecided,therefore,toexaminewhethertheglobalwarmingalarmisan
unusual phenomenon. To do so, we used the structured analogies approach. 39
The structured analogies method involves asking experts to think of similar situations
tothesituationofinterest.Informationisthenobtainedabouttheoutcomeofeachanalogy.
We had previously tested the structured analogies method for forecasting complex
situations involving interactions between parties with conflicting interests, including a
special interest group occupying a public building and demanding taxpayer funding, and
an international crisis over access to water. The research found that a structured search for
and analysis of analogous situations produces forecasts that are much more accurate than
the usual method of asking experts what they think will happen. 40 Other researchers have
subsequently found the method useful for forecasting the outcomes of policy initiatives. 41
With the help of domain experts we have, to date, identified 26 analogous situations. 42
They all began with an allegedly portentous incident or with claims that an apparent trend
was ominous. Searches for evidence supporting each alarm followed, along with calls for
government action. In no case was there recourse to scientific forecasting. The fact that we
were able to identify as many environmental alarm analogies as we did, and the frequency
with which they have occurred in recent times, suggest that they are a common social
phenomenon and that the global warming alarm is not at all unusual. More generally, it