Ken Stewart, 2 January 2014
UPDATE 3 January: BOM has updated it’s time series graph, but not the raw data, which still finishes at 2012! See below.
I have calculated the annual 2013 minimum temperature anomaly for Australia, well before the Bureau of Meteorology.
Not including the 8 sites acknowledged as having anomalous warming due to the Urban Heat Island (UHI) effect, I calculate the straight mean (without area averaging) to be +0.82 C. This puts 2013 as second warmest after 1998, and just ahead of 1973 and 1988.
I expect that the BOM will publish a figure of around +1.2C, and claim 2013 as the warmest on record for minima.
I calculated this by using daily Acorn data for 1910 to 2012 from http://www.bom.gov.au/climate/change/acorn-sat/ , plus daily minima for 2013 for these same sites from Climate Data Online. I used Acorn data from 1961-1990 to recalculate monthly means for each site, and then calculated running centred 31 day means to estimate daily means for the same period.
Then I calculated daily anomalies for each site, and amalgamated these into a straight mean for Australia.
The result is as follows:
I will analyse Fig. 1 in some detail later. But first, how does my calculation stack up against the BOM super computer?
My calculation is in green, BOM in red. As you can see, the match is pretty close, and of course I have not used any area averaging. But you would expect the results to be close, as I have used exactly the same data. You will notice that the major differences occur in years of higher or lower than normal minima. These appear to have become larger in the last 40 years. The official annual figures show greater extremes, as shown above.
I have also calculated trends for the 1910 to 2013 period, and hope that this will persuade you of the futility of using linear trends for temperatures, and that if you cherry pick you can prove just about anything. The next graph is a plot of the continuous running trend from 31 December 2013 all the way back to 1 January 1910. That is, the linear trend through datapoints between any selected date and 31 December 2013.
The vertical axis measures trend in degrees Celsius at particular points in time. Note the rapid fluctuations at the right hand end. I’m sure no one would be silly enough to calculate trends of only a few years’ data.
As the time period increases (moving from right to left) the fluctuations smooth out. Note that Australia has had zero trend in daily minima since 21 July 1997. Interesting, but no predictor of the future.
Moving further back in time, the plot shows the temperature trend increasing until the early 1940s. Up until then the long term trend is fairly stable. Since 1910 the trend is about 1.1C per 104 years. The maximum trend can be calculated from 1922. Therefore, a cheerful cherrypicker can choose whatever time frame they like to produce a linear trend that suits.
Note that the 365 day mean peaked in early November 2013 and has dropped since then. The peak was at +0.94C, which is still below that of 1998 and 2006.
But also note that the rise of about +1.1C over 104 years is by no means steady. There are several sharp rises and falls along the way. Let’s have a closer look at these.
I have shown (starting in 2014) how the minimum temperature record of Australia features a series of sharp step ups, followed by slow declines. I have indicated the start of these periods and the linear trend lines of each one. There may have been one in 1926, and 2013 may (or may not) be the start of another such period. They are more frequent and more pronounced in the past 40 years than in the first 60 years. This appears to show a link to natural climate forces, such as the El Nino- Southern Oscillation.
I will analyse these results further in future posts, and may do the same for maxima as well. (People are interested in maxima because “that’s how hot it is”. I like minima because they tell you more about climate e.g. if they increase faster than maxima this may indicate greenhouse warming.)
Watch for the official 2013 minimum temperature anomaly: probably +1.2C.
Update 3 January:
Here is the official BOM graph to 2013:
and it looks like a bit over
+0.9C +0.94 C, so less than I expected and closer to mine.
Tags: Australia, Australia’s Bureau of Meteorology, bom, Bureau of Meteorology, climate, CO2, minimum temperature anomaly, Surface Air Temperatures, temperature, temperature anomaly, temperature change, temperature records