David Bidstrup: Is it getting hotter in Boulia?...

With summer just around the corner we should be ready for the usual hype about heat waves and “hottest days on record”. These events will be presented as “proof of climate change” so it is worthwhile putting this into perspective. Some readers might have seen a couple of my musings on temperature records for Adelaide and Sydney some time ago but this time I have found a set of records spanning 120 years for the Queensland outback town of Boulia.

With droughts affecting many areas of the country I wanted to see if there was any noticeable change in temperatures in the outback. Boulia provided the longest record I could find. It’s a bit out of the way sitting almost half way between Birdsville and Mt Isa and the country is pretty unforgiving.

The records are sourced from the Bureau of Meteorology and cover the period from 1888 to present. Like most BoM records there are holes in the data. There are 47,783 days in the data record but 2,609 have no temperature assigned leaving 45,174 actual records or about 90% of the total record. In the past I have noticed most BoM records have holes in them.

In this analysis I am looking only at maximum temperatures for two reasons. First, any use of “averages” is meaningless and of no value and secondly I am looking to see if it is getting hotter absolutely rather than relatively.

The analysis looks at the temperature record in nominally 10 year blocks, the differences being the first period from 1888 to 1900 is 12 years and the last from 2010 to 2017 is 7 years.

Each block is analysed to find the maximum temperature ever recorded and the number of days 40 degrees and above. I chose 40 degrees because Boulia is a hot place. The locals would be putting on jumpers when it got down to 30.

The table below shows the analysis.

The first column gives the time period. The second gives the number of days in the period, (remember the holes in the data mean that there will not be even numbers for each period). This is followed by the number of days in the period when the maximum temperature was 40 degrees C or above and that is expressed as a percentage in the next column.

The highest temperature for the period, T max, follows and the mean of Tmax records, (for reasons I will outline below) and the difference between T max and “the mean”.

The last column gives the date when the corresponding T max occurred, (For some reason I cannot get 12 December 1888 to format like the rest – technology!!!!!).

I have also shown the lowest temperature ever recorded at Boulia, -1.4 degrees C.

When the results are graphed it is clear that there has not been any “warming” in Boulia for around 120 years in fact the absolute T max has been fairly steady if a bit in decline – see the blue  trend line on the chart. The days above 40 degrees range from 9% to 18% and “on average” occur 13% of the time so Boulia should expect around 48 days each year above 40 degrees.

On the graph below the absolute T max for each period, (series 1), is shown along with the “mean” of all temperatures for the entire record, (series 2, remember this looks only at maximum temperatures). The variation between T max readings and “the mean” vary up to 16.5 degrees.

When you look at the range of temperatures that the good folk of Boulia have lived within over the 122 years of records it is 49.7 degrees C, from the lowest recorded minimum of -1.4 to the highest maximum of 48.3.

I also used a technique I learned from Dr Darko Butina where the temperatures are analysed to see whether the data fits a bell curve. This is done by calculating the “distance” in standard deviations that each record sits from the mean, and then seeing whether 98% fit within +/- 2 standard deviations. In this case  0.12% lies outside +2 standard deviations and 1.8% outside -2 standard deviations leaving 98.08% within the +/- 2 range. Anyone interested should look at the articles on his site for the details.

I suspect that analysing records for other places in the region like Longreach, Mt Isa, Winton, Bedourie and Birdsville would yield the same results. The problem is finding a record of sufficient time duration. Boulia had 122 years so it was a good one.

My conclusion is that nothing has changed over 122 years. Boulia is always hot but it is not getting “hotter”. The “hottest day on record” was February 7 1915, over 100 years ago.