More on 2009 Global Temperature Data
Posted January 18, 2010
Late Friday night Jim Hansen posted the results of the NASA-GISS analysis of 2009 global temperature data to his Columbia University web site. (The data should be posted on the GISS site later this week). Like NOAA, which I discussed in my previous post, GISS finds that the 2000s was the hottest decade on record, significantly surpassing the 1990s.
The GISS analysis lays to rest claims by climate change deniers that global warming has leveled off in the last decade with this figure showing the five-year and eleven-year running averages. (Note to commenters: genuine skeptics are good for science, but people who consistently and willfully ignore the data are “deniers.”). As I noted previously, the trend in this kind of multi-year average is far more significant than year-to-year changes, particularly with selected starting or ending years, which can radically distort trends.
Looking deeper into the data does reveal some interesting differences between the NASA and NOAA time series, which are most likely due to slightly different approaches to combining all the individual weather station data to obtain a global average. In the NASA-GISS time series 2009 is tied for the second warmest year on record with 1998, whereas NOAA pegs it as tied for 5th. (Both NOAA and NASA rank 2005 as the warmest year, in contrast to the UK Met office which ranks 1998 as slightly warmer.) Hansen’s post estimates that the uncertainty in determining the global average temperature in any one year is 0.025ºC (one standard deviation). Taking this uncertainty into account, 1998, 2002, 2003, 2006 and 2007 are in a virtual tie with 2009 for second warmest in the GISS data series. Similarly, 2002, 2006, 2007 and 2004 could all be considered to be in a statistical tie with 2009 for third warmest in the NOAA dataset.
Again, the bottom line is that the long term trend in the multi-year running mean is far more significant than the rank of any given year. As the figure shows, and as the IPCC concluded, that trend is unequivocal.