In The Weekend Australian I wrote that many false claims are made about the state of our environment on an almost daily basis but because most Australians are illiterate when it comes to science and maths, they are mostly just accepted. My first example was eastern Australian rainfall and the claim that the east coast of Australia has suffered declining rainfall, a claim first made by Sir Nicholas Stern in his influential report to the British Parliament.
I explained that observational data on rainfall for the entire east coast of Australia is available from the Australian Bureau of Meteorology with yearly averages for all the sites back to 1900. But, contrary to the Stern report, this chart does not show declining rainfall; rather, it indicates that rainfall was very low in the early 1900s, that there were some very wet years in the late ’50s and early ’70s, and overall the trend is one of a slight increase in rainfall during the past 107 years, Chart 1.
While I received many emails, phone calls and The Australian published several letters supportive of my analysis, there have been criticisms including the following comment published in The Australian, “A number of scientifically indefensible ways of presenting data are used in the “Case of the warm and fuzzy”. In each case the errors support the author’s conclusions… In two of the six graphs shown the data is fitted to a straight line which is claimed to show an upward trend in rainfall. What would happen if a non-linear fit (eg, quadratic) were used instead of linear fit.. Dennis Matthews, Ironbark, SA.”
When I fit a simple quadratic equation to the data, it also shows an increase in rainfall over the 107 year period, Chart 2.
But there is really no reason to assume that the rainfall data would be represented by any particular equation (linear or quadratic). These trend lines are unlikely to provide any insight into what is likely to happen next year because weather systems, like financial markets, are complex.
But interestingly, even fitting an 11 year moving average shows a trend of increasing, not decreasing, rainfall, Chart 3.
My advice to those trying to interprete data presented graphically would be to use what Professor Harry Roberts, University of Chicago Graduate School of Business, has described as the world’s most powerful statistical analysis tools – your eyes. What can you see from the squiggly line, Chart 4?
Hat tip to MR for the Harry Robert’s advice.