The media has recently been reporting apparently unprecedented heat in Central Australia in the context of human-caused climate change. But is the current heat wave, with extended periods of days above 40oC at Alice Springs really unprecedented? To answer this question it is necessary to examine the data.
There are two sites at Alice Springs for which readily accessible temperature data are available. The first is the Alice Springs Post Office commencing in 1878 and ceasing in 1953; the other is the Alice Springs airport commencing in 1941 and currently the official observing site for Alice Springs. The sites are about 10 km apart; the difference in January monthly mean maximum temperatures between the sites during the period of common observations (1943-53) was 0.2oC with the airport being the warmer of the two.
For the airport site the January monthly mean maximum temperature for all years of record (1942-2012) is 36.2oC. The monthly mean January maximum temperature for all years (1879-1953) at the Post Office is 0.3oC cooler at 35.9oC. The impression is that, when combined, we have a relatively homogeneous maximum temperature record for Alice Springs that spans 134 years.
The airport site is the basis for conclusions being drawn that warming has occurred during the second half of the 20th century. The 95th percentile for the monthly mean maximum temperature data is 39.0oC; five years exceeded this value making the hottest Januarys 1994, 1999, 2004, 2006 and 2008, all in recent decades. The warmest year (2006 with a monthly mean of 40.0oC) started with 12 consecutive days above 40oC and with a subsequent 4 days above 40oC. To date, 2013 is up with these previously hot months having experienced the first 14 days with temperatures reaching 40oC or above. On these data alone one might conclude central Australia has been getting hotter.
The Post Office data, however, show a quite different picture. The 95th percentile value is 39.3oC, or 0.3oC warmer than for the airport. The hottest years from the Post Office record were 1879, 1881, 1887 and 1881, all in the late 19th century. The hottest year in the Post Office record was 1887 and had 11 days above 40oC, a brief respite then another 10 days above 40oC. Taken in isolation the Post Office record would suggest a very warm late 19th century with a cooling trend since.
When we plot the monthly mean data for both sites an extended pattern of cooling followed by warming emerges. Temperatures are now only recovering to the values of the late 19th century.
It is unfortunate that the Australian government has not considered it sufficiently important to digitise and make publicly accessible all of the meteorological records from earlier years. The Bureau of Meteorology website has a range of important statistics about changing climate but most are generated from data subsequent to 1910 and based on a digitised selection of those recorded. As a consequence, statements based on the post-1910 data that suggest an ongoing warming trend are incomplete and likely misleading.
There is fragmentary accessible data (such as the above for Alice Springs) and much anecdotal evidence to suggest that during the late 19th century over central Australia, western New South Wales and South West Queensland the temperatures were as warm as or warmer than for recent decades. Without ready access to the existing earlier meteorological data a faulty picture of a warming Australian climate is portrayed in official statements. However, based on the Alice Springs data, a coming period of cooling cannot be discounted.
Kew, Vic 3101