This aspect of global warming is the one that concerns many people. Will events such as the heat wave that affected much of the USA in 1988 occur more often? Popular images of global warming are of increased extreme conditions and excessive sea level rise. It is these extreme events that would cause the most damage and may have the most impact. Mitchell et al. (1990) detail how the frequency of extreme events might change. Climatic events such as daily average temperature or rainfall form a bell-shaped or normal distribution. This is shown by the solid line in Figure 5.6. Taking temperature as our example, with global warming, mean temperature may increase. The shape of the curve remains unchanged but shifts right. This results in a larger number of hot days but fewer cold days (Figure 5.6a). Alternatively, the mean temperature may stay the same but the shape of the curve might change as shown in Figure 5.6b. In this case the frequency of hot and cold days will increase. Alternatively, the mean and the distribution may change together.
Many problems arise in the analysis in variability. How do you set a baseline, an average climate, from a constantly changing climate? How do you define what is extreme? Changes in meteorological practice and analysis in defining such extremes cause problems when trying to assess variability. As a result, many changes have been found to be due to these inconsistencies. Only tropical cyclones (Hurricanes) in the North Atlantic are thought to have changed and that was a decrease in activity (Nicholls et al., 1996). There is some evidence for less extreme cold temperatures, but no increase in high temperature events, even though mean temperatures have increased. Given the lack of global data on the time-scales needed, no firm conclusions can yet be drawn, but there appears to be no increase in extreme weather events or climate variability at global scales. At regional scales there is some evidence of change, but the change is not uniform with some areas exhibiting greater variability and others lower variability.
One area where there have been distinct changes in variability is in the El Niño Southern Oscillation (ENSO) described in Chapter 2. Since 1976—77 the El Niño has returned far more frequently than previously documented. Instead of the normal 4 to 5 year gap between each El Niño the return period has been about 2 years (Nicholls et al., 1996). There have been few La Niña episodes. As a result there are distinct changes in the precipitation and circulation patterns. This may explain the decrease in hurricane activity, noted earlier, as El Niños are associated with weakened hurricane activity. The El Niño of 1997 was the largest-amplitude event to have taken place since instrumental records began (Jones and Thorncroft, 1998). It would appear that the current ENSO behaviour has not occurred since the instrumental record began about 120 years ago. Whether the ENSO has ever exhibited behaviour similar to the present pattern prior to the instrumental record is unclear (Nicholls et al., 1996). Such studies have to rely on historical documentation of an event, which may well be missing. They also require some means of assessing the
Was this article helpful?