The other main hazards relating to precipitation extremes are hailstorms, snow and ice storms, and drought, with its associated wildfire hazard. Hail develops within thunderstorms, where particles of ice grow by accretion of supercooled liquid drops within the cumulonimbus cloud. Hailstones can grow to large sizes only if they remain in the cloud for some minutes, in the strong updraughts associated with severe thunderstorms. In the United States the main prevalence of hailstorms is in the western Great Plains. Though they generally cause relatively few fatalities, hailstorms are extremely damaging to property (e.g. cars) and to crops. Changnon and Changnon (2000) have carried out an analysis of hailstorm frequency in the United States over the past century. They find large multidecadal fluctuations with considerable regional variations. There are no consistent trends, with most stations reaching their lowest level over the last twenty-year period (1976-95), but some showing a peak at that time. The effects of climate change on hailstorms are not yet predictable, because individual thunderstorms cannot yet be resolved by GCMs.
Extratropical cyclones not only can cause a wind hazard (Chapter 2) and a direct flood hazard (this chapter), but in certain circumstances they can cause other precipitation-related hazards, namely severe snowstorms (including blizzards) and ice storms. The former can cause major disruption of transport, as well as potentially causing river flooding when the snow melts. Ice storms are associated with freezing rain, when rain (or cloud) droplets form thick accretions of ice on surfaces that are below freezing point. Ice storms can cause great damage to power lines and forests, and are a particularly important hazard in the
Great Lakes region of North America. The general predictions of models regarding extratropical cyclones (see Chapter 2) should apply to these hazards, though more detailed regional modelling would be required to assess the effects of climate change, including regional warming, for any particular area.
The final precipitation-related hazard to consider is drought. Smith (1996) noted that a drought may be defined as 'any unusual dry period which results in a shortage of water'. Drought can often be a final trigger for a famine, and has therefore produced some of the world's worst disasters (Smith, 1996). When combined with poor land use practices, drought in dry areas can also lead to desertification (Smith, 1996). However, the drought hazard is outside the scope of this book in that it is a creeping (rather than rapid-onset) hazard, as the expected rainfall fails to occur day after day. Moreover, it is not a simple climate-related hazard, because, though deficiency of rainfall is always the initiator, the drought hazard depends on whether useful water is still available (e.g. in reservoirs and rivers) rather than whether there is simply an unusual lack of rainfall (Smith, 1996). Nevertheless, it is of interest briefly to touch on the immediate cause of drought, which is a prolonged negative extreme of rainfall. The reviews on positive extremes of rainfall, quoted earlier, also include some discussion of drought. Drought shows a large variability, but the areas of the world affected by both drought and excessive wetness are observed to have increased. Modelling studies suggest that the shift in precipitation towards heavier events may also be accompanied by an increase in the number of dry days in some areas. In particular, models predict an increased chance of drought in mid-continental areas in summer, owing to both decreased precipitation and increased temperature and therefore evaporation. The predicted change in the interannual variability of the South Asian monsoon would also increase the chance of drought there, as well as flooding.
Although drought is a creeping hazard, it can lead to the rapid-onset wildfire hazard, as summarised, for example, by Smith (1996). Wildfires tend to occur in regions with a
Mediterranean or continental climate where there is sufficient rain to produce vegetation for fuel, but there is also a dry period in the summer, making the vegetation easy to ignite. Lightning strikes from summer thunderstorms then provide the natural ignition source, though many wildfires are now started by human action, including arson. The wind also plays a major role both in helping to dry out the vegetation before it ignites, and in spreading the fire once it has started. Hence certain weather patterns are often associated with wildfires, such as warm, dry, unstable air with strong surface winds ahead of a cold front. Topography also plays a part, with fire driven upslope spreading particularly rapidly. Major wildfires are extremely common in Australia, where the vegetation and climate are especially conducive, but many other regions, such as southern Europe and the western United States, are also very prone to wildfires. The main hazard is to forestry and to buildings in the fire zone, as well as to the lives of people caught in the fire's path, including firefighters. The effect of climate change on wildfire frequency is complex, operating via weather variables (particularly temperature, relative humidity, wind speed and precipitation, which together determine 'fire weather') as well as the frequency of cloud-to-ground lightning strikes, and the length of the fire season (Flannigan et al, 1998). For example, Flannigan et al (1998) presented evidence that in much of the world's northern boreal forest, despite an increase in temperature there has been a decrease in fire frequency since the end of the Little Ice Age, because of increased precipitation. They predicted a further decrease in fire frequency there under global warming, by using GCM modelling of conducive fire weather, though without considering lightning strike changes. Their study also showed large regional variations in the effects of climate change, with significant increases in the frequency of wildfires predicted for central continental areas, in accord with the drought predictions above. Beer and Williams (1995) have also modelled fire weather using GCMs and predict an increase in wildfires over much of Australia, with slight decreases in relative humidity being the main determining factor. Modelling of fire frequency in the south-western United States by Price and Rind (1994), again using a GCM but including changes in the lightning regime, also predicted an increase in the frequency of fires there under global warming.
The final hazard we will consider briefly is that of temperature extremes. An average global surface temperature rise is the best-known, and arguably the most certain, observation and prediction relating to climate change. However, there has been less research on changes in extremes of temperature. This is despite the fact that unusually hot days in summer or cold days in winter, especially if forming a multiday period (a heatwave or a cold wave), are an important hazard with significant socioeconomic impacts (for example, see Kunkel et al., 1999). Temperature extremes can affect human health and cause many deaths through heat stress or hypothermia, though these are often linked to humidity and wind chill respectively (for example, see Tobin and Montz, 1997, pp. 53-55). Temperature extremes can cause property damage too (e.g. through damage to frozen pipes). Frost is also a well-known hazard for agriculture, and crops can be damaged by prolonged high temperatures as well (Mearns et al., 1984).
Observations of changes in temperature extremes have been reviewed by Karl and Easterling (1999), Harvey (2000) and Easterling et al. (2000a, b). Over the period 1950-93 there were larger increases in mean surface air temperature over land than the global average, and the increase in (mainly night-time) mean minimum temperatures was twice as great as the increase in (mainly daytime) mean maximum temperatures, so that the average diurnal range has been reduced. These asymmetrical changes in the daily mean values are also reflected in changes in extreme temperatures (Table 3.1). One consistent and widespread peaks-over-threshold result is an increase in the number of frost-free days per year, with a longer frost-free season, which has been reported for several regions of the world. Furthermore, for a number of regions that have been examined, the lowest daily minimum temperatures (within certain periods, e.g.
other hydrological and temperature extremes 63
Table 3.1 Summary of analyses of temperature extremes around the world (Source: Easterling et al., 2000a. Reproduced with permission from the American Meteorological Society)
Was this article helpful?