Temperatures in Antarctica

A week or so ago I lambasted the press for trying to portray the collapse of a small portion of an ice shelf on the Antarctic penninsula as evidence of accelerating global warming.  I argued that for decades there has indeed been a warming trend on the Penninsula  (less than 5% of Antarctic land area) but a cooling trend in the rest of the continent.  This implies that the ice shelf collapse either 1) means nothing, as ice shelves do collapse and re-grow fro mtime to time or 2) is an indicator of a local warming anomaly.  In other words, the conditions on the Antarctic Penninsula are not representative of the rest of Antarctica, much less the globe.

Today, Roger Pielke’s blog brings more evidence of how the Antarctic Penninsula is behaving differently from the rest of Antarctica:

Surface snowmelt in Antarctica in 2008, as derived from spaceborne passive microwave observations at 19.35 gigahertz, was 40% below the average of the period 1987–2007. The melting index (MI, a measure of where melting occurred and for how long) in 2008 was the second-smallest value in the 1987–2008 period, with 3,465,625 square kilometers times days (km2 × days) against the average value of 8,407,531 km2 × days (Figure 1a). Melt extent (ME, the extent of the area subject to melting) in 2008 set a new minimum with 297,500 square kilometers, against an average value of approximately 861,812 square kilometers. The 2008 updated melting index and melt extent trends over the whole continent, as derived from a linear regression approach, are –164,487 km2 × days per year (MI) and –11,506 square kilometers per year (ME), respectively.

Negative trends for the period 1987–2008 of the number of melting days (Figure 1b)
over the Antarctic Peninsula are observed at a rate down to –2 days per year for internal areas and about –0.7 days per year for coastal areas. Contrarily, positive trends (up to approximately +0.25 days per year) are observed on part of the Larsen Ice Shelf.

In East Antarctica, positive trends are observed over the Amery, West, Shackleton, and Voyeykov ice shelves, with values of up to +0.7 days per year for Shackleton and +0.8 days per year for Amery. Interestingly, the latter shows negative trends (down to –0.3 days per year) for internal areas but positive values for coastal areas.

Translation:  Most of Antarctica has seen a trend towards less ice melting over the last few decades, with 2008 setting minimum records.  The exception is around the ice shelves, which have seen an opposite trend.  These ice shelves represent a tiny, tiny fraction of the area of Antarctica, and are thought to be melting because the surrounding sea may have warmed a bit (though that is not certain).  Which all goes to show that weather is really complicated and it is totally facile to write that an ice shelf collapse is a sign of accelerating global warming.

  • Scientist

    What’s totally facile is to erroneously claim that ‘98% of Antarctica is cooling’, not just once but many times, and to fail to acknowledge the magnitude of this misstatement even when it is repeatedly pointed out to you.

  • Matt

    I don’t understand the various references to 2008. We’ve only had one quarter of it so far. What am I missing here please?

  • Scientist

    Matt – it’s the start of winter down there now, so the minimum extent of ice cover in 2008 has already been and gone.

  • See the comparison of the Wilkins ice shelf in March 2008 and January 1996 in the image at http://www.appinsys.com/GlobalWarming and the more detailed study of the peninsula at http://www.appinsys.com/GlobalWarming/AntarcticPeninsula.htm.