THERE is concern that if global warming exceeds 2°C per 100 years, there will be catastrophe. There are various ways of anticipating this… I tend to favour the empirical. Indeed, running a ruler over a temperature series can be useful, if we want perspective.
Considering the maximum temperature record available for Lady Elliot Island — an isolated coral cay off the Australian east coast — and after running a trend line through the numbers, we find that we are almost at this tipping point of 2°C, as shown in Figure 1.
There has been a consistent annual temperature rise of 0.0195°C since 1940, which would translate into a rise 1.95°C per hundred years.
The situation appears even worse if we just consider temperatures from 1960. Then the temperature trend for Lady Elliot is more than 2°C per hundred years, as shown in Figure 2.
The mean annual maximum temperature series from the nearby Sandy Cape lighthouse has a very similar pattern for this period from 1960. There is more inter-annual variation in the temperature series from nearby Gayndah, which is typical of land series that will be more affected by droughts and floods.
I’ve spent many years poring over temperature data from a diversity of locations across Australia, and what I’ve found is that when such records are extended back in time — even just a few decades — the overall temperature trend is quite different and ‘the catastrophe’ disappears.
The record for Lady Elliot Island only starts on 1 July 1939. To understand what temperatures were like in this region before then, we need to consider temperature measurements from neighbouring stations that begin in the late 1800s.
There are several weather stations within a 300 km radius of Lady Elliot Island that have long temperature records.
The temperature record as measured at the Gayndah Post Office (number 39039) actually begins in June 1893, and the record for Bundaberg (number 39015) in 1892, as shown in Figure 3. There is a single breakpoint in the early record for each of these two locations, marked by a circle when a Stevenson screen was installed.
Considering the longer records within the 300 km radius, maximum temperatures have sometimes fluctuated by more than 2°C within a few years. This is evident from the detail in the individual series in Figure 3. The more than 2°C variation in temperatures as measured at Gayndah between 1900 and 1904, for example, is more than the overall warming trend at Lady Elliot of 1.95°C per hundred years since 1940.
Maximum temperatures for Gayndah and Bundaberg spiked in 1915, as they did for all the other stations recording within a 300 km radius of Lady Elliot Island at that time.
So, we can perhaps assume it was also a relatively hot year at Lady Elliot Island … back in 1915. Perhaps it was as hot back then as it is now, which is quite hot considering the longer records. Then again the spike may have been a consequence of the land drought, and places like Sandy Cape Lighthouse and Lady Elliott less affected. There is no mean annual maximum value for Sandy Cape Light House (number 39085) for 1915 because of missing values … but these could be infilled (using regression and/or an artificial neural network), to better estimate the likely temperature history for this region over land and also sea.
It can be difficult separating out the individual series in Figure 3. The squiggly lines are a representation of individual annual mean maximum temperatures. In Figure 4, I plot these individual values for just the Gayndah Post Office (number 39039) from 1900 (after the breakpoint, only after the Stevenson screen was installed) and for Lady Elliot (number 39059) from the beginning. Both series are plotted with trend lines.
The overall temperature trend for Gayndah Post Office, considering this longer record, is just 0.38°C per hundred years, as shown in Figure 4. This contrasts significantly with the value of 2.58°C per hundred years when we plot the same values for Gayndah but only for the interval from 1960 to 2009, as shown in Figure 2.
Whenever I take the time to run a ruler over temperature series for any particular region and considering a long enough time period not artificially warmed by a combination of Urban Heat Islands (UHI) and homogenisation, I have trouble finding catastrophe. Rather I’m inspired by the extent of the available data and the degree of synchrony between the series that tend, if anything, to be simply moving sideways.
This is the third post in my ‘GAT in the Hat’ series, which began when Huck suggesting we get on and develop a simple temperature index based on a good sample of well-sited stations. Jaco Vlok and I are working our way around Australia with such an index in mind … we are working anti-clockwise from Brisbane.
Thanks to the Australian Bureau of Meteorology for making all this data available at http://www.bom.gov.au/climate/data/
The feature image (very top of this blog post) is a plot of all the available maximum temperature data (from 59 weather stations) within a 300 km radius of Lady Elliot as 12-month moving averages, with the data compiled and plotted by Jaco Vlok.
This post was updated at midnight on 3rd August, with figures added, the picture changed, and the title modified …