Winter 2006-07

Why so warm?
by Philip Eden

One of the main difficulties when analysing our climate is to distinguish between underlying trends and short-term fluctuations. The last twenty years present an excellent illustration of this.

We all know from experience that winter months dominated by easterly winds are much colder than those dominated by westerlies. We can easily quantify this: before the accelerated warming trend evident since 1989, the five most -westerly- Januarys had a mean Central England Temperature (CET) of 6.2°C while the five most 'weasterly' Januarys had a CET of 0.4°C. The same analysis for February gave 6.2°C for the westerly months and -0.5°C for the easterly months, and for March respectively 7.6°C and 4.2°C.

In the UK the winters of the 1990s were more than 1°C warmer on average than the winters of the 1960s. However, westerly winds were much more dominant during the more recent decade. Averaging the barometric pressure pattern for each month for the two contrasting periods shows that the Januarys of the 1990s were 2.5 times more 'westerly' than those of the 1960s, February was 4 times more westerly, and March 3.5 times more westerly.

It is quite reasonable, therefore, for someone to ask whether some, most, or all of Britain's winter warming since 1989 has been caused by this change in weather pattern. We can now offer some tentative answers, because there has been a sharp downturn in the westerliness of winters since 2000. The figures show that the late-winters and early-springs of 2001-2007 exhibited a 45 per cent decline in the strength of westerlies compared with the 1990s, which, applying the results obtained from the pre-1989 analysis, should have produced a drop in temperature of 1.5°C.

Looking at the individual months shows clearly what is happening. Comparing the periods 1989-2000 and 2001-2007, January has experienced a 10 per cent decline in westerlies, yet the CET has risen by 0.3°C February is 60 per cent less westerly but the CET is a mere 0.2°C lower; March (up to 2006) shows a 70 per cent drop in westerliness associated with a drop of 0.5°C in the CET.

So, yes, 'some' of the warmth of the 1990s was merely a decadal fluctuation, but there was also a strong underlying upward trend in temperature.