"The art of prophecy is very difficult, especially about the future." - Mark Twain

Has the Antarctic continent become significantly warmer in recent years? 

Are there any reliable, long-term temperature records available for this region?

Do historical temperature records give us any indication of probable future temperature trends?

Are the snow and ice deposits in Antarctica likely to melt and result in significant worldwide rises in sea level?

Some of these questions are answered below.   Please send me your comments.

Long-term air temperature records at 13 stations in Antarctica have been extracted from the website of KNMI (Royal Dutch Meteorological Institute).  This is all raw data (GHCN), without any adjustments ever having been made (as far as I am aware).  Only stations with records over a period of at least 50 years, with largely complete and post-2005 records were selected.  Eleven of the stations are located in coastal areas of the continent and of these three are located on the more northerly Antarctic Peninsula (south of South America).  Two stations are located at higher, inland locations including one station at the South Pole.

The latest data that I could locate from the KNMI database were for July 2011.  For the three Australian operated stations later data were obtained from the Australian Bureau of Meteorology website.

I have not been able to make any quality checks on the data.  It is possible that there have been some changes in station locations and instrumentation over the periods of record, but comparisons with other adjoining stations would be expected to identify any inconsistencies.

Graphical plots are presented below of annual mean temperatures at each of the 13 stations.  [ Annual mean temperatures are the average of the mean daily temperatures over each year (the mean daily temperatures are the average of the observed daily maximum and minimum temperatures). ]   

In addition, the mean monthly temperatures for the warmest and coldest months (based on long-term averages) at each station are plotted.  The coldest months (on average) are always July or August while the warmest month (on average) is January (except for the two inland stations that are slightly warmer in December). 

A trend line was fitted through each of the sets of data.

MY CONCLUSIONS:

1.  Apart from at the three stations located on the Antarctic Peninsula, none of the records at the other ten stations show any trend to either increase or decrease in temperature over the past 50 years.  This conclusion applies to the mean annual temperature, the mean temperature during the warmest summer month (December or January) and the mean temperature during the coldest winter month (July or August).

2.  The three stations located on the Antarctic Peninsula all have increasing temperatures over the past 50-60 years, although there are significant differences in magnitude.  The mean annual temperature increase varies from about 1ºC (at Esperanza) to about 4ºC at Rothera Point.  The summer temperature increase varies from about zero at Esperanza to about 6ºC at Faraday/Vernadsky.  The winter temperature increase varies from about 0.5ºC at Rothera Point to about 2ºC at Esperanza.

3.  I have no concrete explanation for the above results for stations on the Antarctic Peninsula.  However, it seems possible that the three stations on the Antarctic Peninsula are affected by ocean currents moving between the South Atlantic and Pacific oceans, while the other stations would not be so affected.

4.  The zero temperature trends for the main regions of the Antarctic continent are interesting and possibly significant.  In most regions of the world there have been a well-defined, but not alarming, increase (of less than 1ºC/century) in mean annual temperature over the past 100-150 years but this trend does not exist for the main land mass of Antarctica.

UPDATE (7 October 2013):   A more detailed analysis of the temperature trends at Mawson may be seen at:  http://members.westnet.com.au/brigun/mawson-temps.html

Swoya (Japan), Antarctica      40 complete years of data, 1957-2011