Ahead of the upcoming IPCC report into global climate, and climate change, the news agenda seems to been largely dominated by stories asking why global warming has paused for the last 15 years (see the BBC, the BBC again, the torygraph and the NZ herald among countless other examples).
A substantial part of this seems to be the repeat of familiar claims that 1998 was the hottest year on global record, and if global warming scientists were right there is no way that we should not have a seen a hotter year during the past 15 years. Hence, climate change has paused, the models and data suggesting that human fossil fuel emissions were to blame for late 20th century warming were wrong, and that consequently any argument for restricting emissions in future are null and void.
Which of course ought to lead to the question, who says that 1998 was the hottest year on record? Well the answer to this is somewhat complicated, but also somewhat revealing. It aint NASA, who run GISStemp (the Goddard Institute for Space Studies Surface Temperature Analysis) who have 2010 as the hottest year on record followed by 2005, with 9 of the 10 hottest years occurring after the year 2000 (with 1998 as the only pre-2000 year in that list). It also isn’t NOAA (the US National Oceanic and Atmospheric Administration) who compile a global temperature record at the National Climactic Data Centre (NCDC), whose data again places 2010 as the hottest year on record, followed by 2005, with 1998 in third, and 9 of the hottest 10 years on record occurring after the year 2000 (ie after global warming has allegedly paused). Which leaves the UK Met Office’s Hadley Centre and and University of East Anglia’s Climactic Research Unit record HadCRU. The CRU is of course the unit who were the subject of the Climategate faux controversy where sceptics hacked emails, published some excepts from private correspondence out of context claiming fraud and data manipulation generating global headlines, and which were subsequently found by numerous independent investigations to have found no evidence of wrongdoing. The latest version of this temperature series is HadCRUT4v which again shows that 2010 was the hottest year on record, followed by 2005, followed by 1998.
So where does the 1998 was the hottest year claim come from? Well, HadCRUT4v is the latest, and most accurate temperature record maintained by the Met Office and CRU (for a detailed explanation of what’s changed look here). If we ignore that and instead use their previous version, HadCRU3v, then, and only then does 1998 appear to be the warmest year on record. So why did this old record suggest a different year to the NASA and NCDC records (and indeed the latest version of the CRU record)? Well the main reason for this was the different methods used to generate global temperatures. Of course none of these institutions are able to measure the temperature in every place in the world, they use stations in various locations, and the places where there tend to be the fewest stations tend to be the polar regions (where there also tend to be the fewest people). And one of the things we know quite well, is that the Arctic has been the fastest warming region on the planet. Whereas GISStemp interpolates values between measured locations in the Arctic, HadCRU3v left them blank as unknown, which introduced a cold bias into their dataset compared with the others, and explaining why it has been replaced by a dataset which features a greater number of stations and which correlates much more strongly with the other datasets.
So the ‘pause’ in climate change is something that only exists if you exclusively look at a now obsolete and known to be biased dataset generated by a group who those using this data have previously claimed to be frauds. And decide to ignore that 1998 was in any case a super El-nino which had a dramatic short term effect on global weather – hence the other 9 of the 10 hottest years on record all occurring since the year 2000. If you used 1997 or 1999 as start dates there wouldn’t appear to be any pause in any dataset (outdated or otherwise), but cherry-picking the year when specific short-term conditions made things abnormally hot added to cherry-picking a now obsolete dataset allows sceptics to make the global warming has paused argument (see this excellent skeptical science post for details on cherry-picking)
So why are so many mainstream media outlets focussing upon this as the main story in the lead up to the IPCC report? Probably because it’s a more sensationalist and conflict-driven story than one which reads science has been slowing progressing, turning a 90% confidence in predictions in 2007 into a 95% confidence by 2013, allied with a big PR drive from a number of the main players in the climate denial industry.