Tuesday, January 02, 2007

Meteorological bias

We have a curious bias for the statistics of weather which happens to fall within a calendar year or a calendar month. For example, in a post last week I pointed out that although July 2006 was the equal hottest July ever recorded in the UK, the hottest ever recorded 30-day period actually occurred from the middle of July 1995 to the middle of August 1995. Because this period of heat happened not to fall within a calendar month, it is not enshrined in our meteorological statistics. Similarly, newspaper headlines have claimed that 2006 was the hottest ever UK year on record, but in fact, as Philip Eden points out, the hottest ever recorded 12-month period was actually that between November 1994 and October 1995.

Global warming, if it exists, exists as a trend, not as something which can be verified or falsified by the weather in one particular year; just as the heat in 1995 did not prove the existence of global warming, nor did the heat in 2006. Furthermore, whilst the global trend is undeniably upwards, it should be noted, as a curiosity, that the average annual temperature has been falling in Australia over the past couple of years.

No comments: