Heat waves of Gabo (and nearby)

Reader advisory

The linked documents are pretty lengthy (around 30 pages I think) and mainly consist of my rambling through piles of numbers. Several charts. You have been warned

Following from my previous post about Heatwaves (or as I call them “Hot Spells” and “Very Hot Spells”) for Mallacoota I wondered whether the longer period of observations for Gabo Island Lighthouse might give a more definitive answer.  Looking into this proposition took me a lot of time, exploring many dead-end burrows in a rabbit warren: at some points I turned so many times I nearly disappeared to “the place where the sun don’t shine” (see second and third examples in this). 

Examining the 110 year series for Gabo Island generates a series that looks at a glance to show an increasing number of Very Hot Spell Days in the warmer season.  However a very rough trend analysis does not show statistical significance.

I noted that there were quite a lot of missing values - particularly for minimum temperatures and particularly in later years.  I explored the idea of using data for Green Cape to impute values where they were missing from the Gabo series.  After considerable fiddling with the Green Cape data it seemed that there might be some value in imputing missing values for the final 20 years of the Gabo series.

Making these imputations increased, by about 9% the number of Very Hot Spell Days in those years but not enough to make a statistically significant difference..

My big ticket outcome is that I can’t say there is a significant increase in the number of Very Hot Spell days on Gabo Island.  Possibly a meteorologist may know of techniques that would demonstrate that, but possibly they would find flaws in my approach and dismiss it completely.