In spite of a bit of a cold snap in parts of the northern hemisphere recently (we’ll come back to that) it’s been hot just about everywhere in 2009. In Australia 2009 has been the second warmest year on record after 2005, and the hottest decade:
Figure 1: Annual and decadal mean temperature anomalies for Australia (compared with 1961-90 average)
The Australian pattern shows a consistent warming from the 1940s.
In 2009 was especially hot in NSW and SA as this map shows:
Figure 2: 2009 mean temperatures compared against historical temperature records
Please note that this was a day/night mean.
Australia, Victoria, South Australia and NSW all recorded their warmest July-December periods ever. The following map shows the maximum anomaly:
Figure 3: Maximum temperature anomaly for July-December 2010
Also Australia has been increasingly getting more record hot days.
While both high temperature records and low temperature records continue to be set in Australia, we’ve seen over the last 50 years high records have increasingly outnumbered low records and over the last 10-12 years, high temperature records have been broken at a rate two to three times higher than low temperature records are.
On a planetary level, NASA GISS have declared 2009 as the second warmest year on record in statistical tie with 1998, 2002, 2003, 2006, and 2007. The hottest year on the GISS record was 2005. What, 2005 the hottest and not 1998? Well that’s the way GISS see it. In this RealClimate post James Hansen gives a detailed explanation as to why. In brief, GISS estimate the areas where there is no station coverage by using the nearest stations. HadCRUT simply leave these areas out.
GISS point out that either way involves making a judgement. The HadCRUT methodology assumes that the blank spaces are at the global average. The GISS approach seems to me superior, especially since the biggest gaps are where the warming is strong, in the polar regions.
The NASA GISS 1880 to 2009 land-ocean temperature index looks like this:
Figure 4: NASA GISS land-ocean temperature index
Tamino at Open Mind in his discussion of the story has the same graph without the clutter of the 5-year mean and the green error bar, so that you can see clearly the values since 1998:
Figure 5: Open Mind version of NASA GISS land-ocean temperature index
The Hansen post explains that if you use a 5-year running mean it suppresses most of the noise generated by ENSO (El Niño and La Niña). Using an 11-year mean minimizes the effect the sunspot cycle. So we get this:
Figure 6: 5?year and 11?year running mean temperatures in the GISS analysis of (a) global and (b) hemispheric surface temperature change against a base period of 1951?1980
In the hemispheric record two differences stand out. Firstly, warming is less in the south because there is more ocean there. Secondly, there is a much less accentuated mid-century dip in the Southern Hemisphere.
You will notice right at the end a slight hook or flattening which desperate sceptics/denialists will still perhaps go ape over. In an earlier post Tamino at Open Mind used statistical analysis techniques to eliminate the ENSO factor along with the aerosol effect of major volcanic eruptions. The result is this:
Figure 7: Adjusted land-ocean temperature index
This casts a whole new light on the years 1998 plus. You might say that a consistent underlying warming pattern has been revealed.
Tamino did that post late last year when the December temperature was not available. As this graph shows, December (the red dot in the bottom right) was perhaps slightly above the yearly average, which would have made no appreciable difference to the adjusted graph in Figure 7:
Figure 8: Annual mean surface temperature anomaly
That red dot, however, contained the cold snap across much of the northern hemisphere, which looked like this:
Figure 9: Global map of December 2009 temperature anomaly
According to Hansen’s post, the December 2009 event was tied up with the Arctic Oscillation (AO) Index the monthly values of which are represented as follows:
Figure 10: Arctic Oscillation (AO) Index
As you can see nothing like December 2009 has happened since 1977. What happened is that there was a cold air outbreak from the Arctic to the mid-latitudes, with the cold Arctic air being replaced by warmer air from the south. So while the temperature anomaly reached as much as -8C in Siberia it was +7C in Alaska. Greenland was also much warmer than usual. So for global temperatures it means squat, nix, nothing as the red dot for December in Figure 8 above shows.
It seems that there has been a tendency over the past few decades for the middle latitude jet stream to blow strongly and consistently from west to east, thus keeping cold Arctic air locked in the polar region. December 2009 was highly anomalous and probably things will return to the previous pattern, but we’ll just have to wait and see.
So the bottom line is that the Arctic Oscillation Index anomaly “caused” the big freeze in December 2009. As to what caused the AO anomaly, it’s a matter of stuff happens. One month doesn’t mean much.
Elsewhere Joe Romm at Climate Progress had a post on 2009 as you’d expect.
PS For a long time I had maroon borders around the images I post. Then last year they disappeared. This week they came back. I do wish they’d go away!
Update: Here’s an image from the IPCC AR4 WGI Summary for Policy Makers – pdf showing models run with (a) natural forcings only, (b) natural plus anthropogenic forcings, and (c) actual observations (decadal averages):
The blue bands represent the natural forcings of solar activity and volcanoes, while the red represents natural plus anthropogenic forcings. The black line represents observations.