Why Tilo Reber is Wrong

(c) 2008 by Barton Paul Levenson

The idea that "there was no warming from 1998 to 2007, therefore global warming is wrong" continues to float through the blogosphere. There are two main reasons why it's wrong.

1. It's wrong because ten years is too small to talk about climate trends. The World Meteorological Organization (WMO) defines climate as "mean regional or global weather over a period of 30 years or more."

2. More importantly, it's wrong because the start date of 1998 is cherry-picked, and the recent 1998-2007 end points are more so. What the denialists do here is to pick out a part of the curve that starts with a strong El Niño year (the strongest El Niño on record, in fact), and ends with a La Niña year.

El Niño and La Niña are opposite points in a climate oscillation, an interplay of ocean and air temperatures in the tropical Pacific. El Niño years tend to be anomalously hot. La Niña years tend to be anomalously cool. Over the long run they even out. No net change.

What the denialists are doing, therefore, is to pick out part of the curve of temperature with time which seems to support their point of view, and ignore the rest. This is called "cherry-picking," and will get you a flunking grade in any introductory data analysis course in the world.

Which brings me to Tilo Reber, a frequent poster on climate blogs including Tim Lambert's "deltoid." He insists that the ten years 1998-2007 proves that anthropogenic global warming is wrong, and says so in extremely hostile language. I'll give this much to denialists like Lindzen or Ball or Essenhigh -- they're polite. Tilo isn't, not by a long shot. But let me quote a key post of his where he repeats his usual line about 1998-2007:

If we look at the entire period of my chart with no warming, there is no way to conclude that the only reason that it shows no warming is because of the El Nino at the beginning and the La Nina at the end.

Let's just check that. Mr. Reber doesn't believe in the NASA GISS temperature anomalies. He calls them a lie, blames James Hansen for making up or distorting data, etc., etc. He likes to use the Hadley Centre temperature record.

Now, over the long run, the Hadley Centre and NASA GISS records correlate very closely indeed. They are, after all, making use of many of the same temperature stations and sea temperature readings to create their estimates. But let's ignore that and go with Hadley.

Here are the Hadley Centre temperature anomalies, in degrees Kelvin, for the years 1998-2007:


Do a linear regression of anomalies against time and the slope is positive, but statistically insignificant at p < 0.64. For a significant relationship you want to see p < 0.05 or less.

Now, let's take out 1998 and 2007. Now the slope is 0.024583 and is highly significant at p < 0.035 despite the fact that the sample size was reduced by 20%. How about that.

So actually doing the math proves that including a starting El Nino year and an ending La Nina year DOES make a difference. 1998-2007 is a classic example of a cherry pick, a fallacy of composition, and anthropogenic global warming has not gone away. So Mr. Reber, and anyone else who obsesses over 1998-2007 as some kind of proof that AGW is a failed theory, is WRONG.

Page created:12/15/2009
Last modified:  02/13/2011