It’s just us.

There is a lot of ado about what the weather has been like since 1998. According to some measurements, 1998 was the warmest year on record (see, e.g., here and here). According to these data, 2005 was the second warmest year on record (by vary narrow margins). According to other data sets, 2005 was the warmest year and 1998 was the second warmest (see, e.g., here). What’s not in dispute is the reason why 1998 was exceptionally hot: it was the same year that there was the strongest El Niño of the 20th century. (El Niño is a weather phenomenon with the ocean and atmosphere linked to periodic warming and La Niña is something similar linked to periodic cooling. See here for more information.)

So why the discrepancies between the data sets? The folks of RealClimate give an explanation here. Essentially, HadCRUT 3V data provided by the Climate Research Unit, which suggests 1998 was the warmest year, has significant gaps in its coverage, particularly over the Arctic where warming has been the most pronounced. When you’re missing data from where temperatures have been increasing the most, your data are going to be skewed. So that data are missing an important data source, which is the Arctic. The scientific literature has made note of this and explains why there are some discrepancies as far as which year was the warmest on record.

The septics (a term I borrow from William M. Connolley — see explanation here), however, maintain that 1998 was definitely the hottest year on record so therefore “global warming has stopped for the past 11 years.” This is such a favorite of theirs that it’s gotten into the mainstream media, which, after all, is interested in sensationalism.

Alas, Steven Levitt and Stephen Dubner fall for this specious argument in their new book, SuperFreakonomics. I was a big fan of their work in the first book, Freakonomics, which is why I am disappointed to see they’ve ventured into an area neither of them are competently familiar with and get their facts very messed up. The chapter in question can be read here. For some critiques, see here (Dr. Connolley’s) and here (Dr. Romm’s).

Like I said though, it’s a popular argument. For example, in an argument about a post over at the SCSU Scholars blog (click comments), the specious argument came up multiple times as a means to disprove anthropogenic global warming. How can it be, the septics ask, that CO2 has been rising but temperatures have not since 1998 (again, assuming 2005 was not the warmest year because that would hurt their argument)? It’s a classic case of cherry picking data. If this year isn’t the warmest year, the septic will pick the previous record and ask why the Earth is cooling.

Sorry, it doesn’t work that way. First of all, climate is considered weather averaged over a 30-year period. Not 11 years, but 30 years (as a general rule). So, yes, they’re cherry picking. And it’s not like this is anything abnormal. Year-to-year variation is to be expected (and predicted in the models, contrary to some claims). When we say that temperature will increase in the long term, no one is claiming that they will go up in a straight line, yet this is what the septics are expecting when they make this argument. In reality, however, long-term upward trends are observable even when there are year-to-year natural variations. The IPCC predicts .11 to .64 °C temperature increases per decade during the 21st century. This is consistent with the past decade:


The graph above is annual global mean surface temperature (land and ocean) anomalies in degrees centigrade with a 1951-1980 base using this data set provided by NASA’s Goddard Institute for Space Studies. In more ordinary language, it’s a graph of temperature changes from 1980 until 2008. The black line is the 25-year trend line. The two shorter lines are 10-year trend lines. The most recent decadal trend (1999-2008) is the yellow one showing a 0.19 ºC per decade trend, consistent with the IPCC’s range.

So what does 1998 tell us about the overall trend in temperatures? Not a lot. If anything, it shows that temperatures have been steadily increasing to new extremes since the Industrial Revolution, when humans began to emit greenhouse gases into the atmosphere at massive and unprecedented levels. What we have seen since 1998 is consistent with the anthropogenic model of global warming (see, e.g., my graph above, here, here, and here).