As someone who goes through politics news often, the subject of climate change comes up quite often: how bad it is, does the data fit, and so forth. But even on a story about the Liberal Democrats as a whole, some people, often right-wing, will attack the party on its green credentials. This stems from a belief in climate change is a “hoax” of sorts. Like this comment on a Telegraph article:
Yeah, Chris Hunhe off to Cancun, tilting at windmills, saving the planet.
Didn’t you know that we’ve had no warming since about 1998?
The Man-Made global warming idea is just a scam.
And you wonder why the LibDems are held in contempt.
As a mathematician, this makes me absolutely cringe when I read this. This stems from the fact that in several graphs, 1998 is apparently the “hottest year on record” (when in fact, 2005 surpassed it). But the flaw in the argument? 1998 is an outlier — a freak result of the data collected.
Let’s look at it this way. Suppose Liverpool F.C. have, for the past ten years, slowly climbed from languishing in around 17th in the Premier League to a comfortable mid-table position, let’s say 11th. Then one season, perhaps because of sheer luck, they finished second. If, over the next ten years, they become stronger and contend for fourth or fifth place, do we say that “they weren’t as good as they were ten years ago”? No. Likewise, you don’t criticise a batsman in cricket for never surpassing his fluke double century.
Statisticians have a way of seeing the trend of data: a regression line. It’s a relatively simple concept that is first taught to children when they start high school, as a “line of best fit” — or in other terms, linear regression. Other more complicated data sets and trends have different kinds of regression lines, including exponential curves and logarithmic curves. And it’s easy to process data sets and fit trend lines, even easier with computer applications like Excel.
I’m using this NASA data-set with three-month averages, from 1976 to 2010. We can see just by looking at the raw numbers that there’s a steady rise. And Excel verifies this, by giving us several regression lines for each season:
This of course, is a very simplistic way of looking at things. Other methods include “moving averages”, or more complicated polynomial regressions.
So why was 1998 so hot? According to New Scientist, who touched on the statistical reason very briefly, it was due to a particularly strong El Niño, a feature of the South Pacific climate that occurs most Christmases which results in hotter, drier weather. While my knowledge of climatology is limited — indeed, I’m doing this solely from a mathematical perspective — the crux of the matter is that several quite probable events happened in 1998 that made it jut out from the norm. So global warming didn’t really stop in 1998; indeed, it’s debatable on whether it’s stopped at all.