|Working Group I: The Scientific Basis|
|Other reports in this collection|
2.7.3 Is There Evidence for Changes in Extreme Weather or
In this section we assess changes in the intensity and frequency of various weather phenomena. One aspect of change that is important, but which is beyond the analysis of present records, relates to changes in the tracks of storms. Severe storms are often rare, so the analysis of large areas and long lengths of homo-geneous storm records are required to assess changes. So far this combination of data is not available.
This section updates the information provided in the SAR regarding changes in tropical cyclones across various ocean basins and those affecting the nearby continents. As reported in the SAR, a part of the multi-decadal trend of tropical cyclones occurring in the Australian region (105° to 160°E) is likely to be artificial, as the forecasters in the region no longer classify some weak (>990 hPa central pressure) systems as “cyclones” (Nicholls et al., 1998). By considering only the moderate and intense tropical cyclones (central pressure 990 hPa), this artificial trend is eliminated. The remaining moderate and strong tropical cyclones reveal a numerical decline since the late 1980s, but the trend is not statistically significant. Similarly, the trend in intense tropical cyclones (minimum central pressure below 970 hPa) is not significantly different from zero. Nicholls et al. (1998) attributed the decrease in moderate cyclones to more frequent occurrences of El Niño during the 1980s and 1990s. However, a weak trend in the intense tropical cyclones implies that while ENSO modulates the total frequency of cyclones in the region, other factors must be more important in regulating their intensity. For example, new work by Higgins and Shi (2000) and Maloney and Hartmann (2001) show that 30 to 80 day Madden-Julian oscillations modulate tropical cyclone activity.
As reported in the SAR, the north-east sub-tropical Pacific has experienced a significant upward trend in tropical cyclone frequency in the short period examined, but additional data since that time show no appreciable trend. There is no appreciable long-term variation of the total number of tropical storm strength cyclones observed in the north Indian, south-west Indian and south-west Pacific Oceans east of 160°E. (Neumann, 1993; Lander and Guard, 1998). For the north-west sub-tropical Pacific basin, Chan and Shi (1996) found that the frequency of typhoons and the total number of tropical storms and typhoons have been more variable since about 1980. There was an increase from 1981 to 1994, which was preceded by a nearly identical magnitude of decrease from about 1960 to 1980. No analysis has been done on the frequency of intense typhoons (having winds of at least 50 m/s) due to an overestimation of the intensity of such storms in the 1950s and 1960s (Black, 1993).
There has been an extensive analysis of the North Atlantic basin for the entire basin back to 1944, and also for the United States landfall tropical storms and hurricanes back to 1899. The all-basin data, however, have been affected by a bias in the measurement of strong hurricanes. This bias has been removed in an approximate way to provide estimates of the true occurrence of intense (or major) hurricanes since 1944 in the North Atlantic (Landsea, 1993). Earlier events lack reliable data on the strong inner core of the hurricanes. The United States record of landfall frequency and intensity of hurricanes is very reliable because of the availability of central pressure measurements at landfall (Jarrell et al., 1992). Both of these data sets continue to show considerable inter-decadal variability, but no significant long-term trends (Figure 2.37, from Landsea et al., 1999). Active years occurred from the late 1940s to the mid-1960s, quiet years occurred from the 1970s to the early 1990s, and then there was a shift again to active conditions from 1995 to 1999. Concurrent with these frequency changes, there have been periods with a strong mean intensity of the North Atlantic tropical cyclones (mid-1940s to the 1960s and 1995 to 1999) and a weak intensity (1970s to early 1990s). There has been no significant change in the peak intensity reached by the strongest hurricane each year (Landsea et al., 1996). As might be anticipated, there is a close correspondence between the intensity of hurricanes in the North Atlantic and those making landfall in the United States (Figure 2.37).
Using historical records, Fernandez-Partagas and Diaz (1996) estimated that overall Atlantic tropical storm and hurricane activity for the years 1851 to 1890 was 12% lower than the corresponding forty year period of 1951 to 1990, although little can be said regarding the intense hurricanes. They based this assessment upon a constant ratio of USA landfalling tropical cyclones to all-basin activity, which is likely to be valid for multi-decadal time-scales. However, this also assumes that Fernandez-Partagas and Diaz were able to uncover all USA landfalling tropical cyclones back to 1851, which may be more questionable.
Extra-tropical cyclones are baroclinic low pressure systems that occur throughout the mid-latitudes of both hemispheres. Their potential for causing property damage, particularly as winter storms, is well documented, where the main interest is in wind and wind-generated waves. In place of direct wind measurements, which suffer from lack of consistency of instrumentation, methodology and exposure, values based on SLP gradients have been derived which are more reliable for discerning long-term changes. Over the oceans, the additional measurements of wave heights and tide gauge measurements provide additional ways of indirectly evaluating changes in extra-tropical storm strength and frequency (see Chapter 11, Section 11.3.3). Global analyses of changes in extra-tropical storm frequency and intensity have not been attempted, but there have been several large-scale studies. Jones et al. (1999c) developed a gale index of geostrophic flow and vorticity over the UK for the period 1881 to 1997. This revealed an increase in the number of severe gale days over the UK since the 1960s, but no long-term increase when considering the century period. Serreze et al. (1997) found increases in cold season cyclones in the Arctic region for the period 1966 to 1993. Angel and Isard (1998) found significant increases in strong cyclones (<993 mb) in the Great Lakes region from 1900 to 1990 during the cold season. Graham and Diaz (2001) find evidence for increases in strong cyclones over the Pacific Ocean between 25 and 40°N since 1948 and link the increase to increasing sea surface temperatures in the western Tropical Pacific. Alexandersson et al. (1998, 2000) similarly studied extreme geostrophic wind events in the north-western European area based on homogenised observations during the period 1881 to 1998. These studies revealed an increase in the number of extreme wind events around and to the north of the North Sea. The WASA group (1998) similarly investigated the storm related sea level variations at gauge stations in the south-eastern part of the North Sea. They found no long-term trend during the last 100 years, but a clear rise since a minimum of storminess in the 1960s, which is consistent with the rise in extreme geostrophic wind found by Jones et al. (1999c). This increase is also consistent with changes in the NAO (Figure 2.30). Some analyses have focused on hemispheric changes in cyclone activity. Lambert (1996) analysed gridded SLP over both the North Atlantic and North Pacific Oceans for the period 1891 to 1991. He found a significant increase in intense extra-tropical storms, especially over the last two decades of his analysis, but the data were not completely homogenised. Simmonds and Keay (2000) used data from 1958 to 1997 in the Southern Hemisphere and found an increase in cyclone activity through 1972 before decreasing through 1997 with strong decreases during the 1990s.
Hourly values of water levels provide a unique record of tropical and extra-tropical storms where stations exist. Zhang et al. (1997a) have analysed century-long records along the East Coast of the United States. They calculated several different measures of storm severity, but did not find any long-term trends. On the other hand, they did find that the effect of sea level rise over the last century has exacerbated the beach erosion and flooding from modern storms that would have been less damaging a century ago.
Another proxy for cyclone intensity is wave height (see Chapter 11, Section 11.3.3). Several studies report increased wave height over the past three decades in the North Atlantic (approximately 2.5 cm/yr) and in coastal areas, though no longer-term trends were evident (Carter and Draper, 1988; Bacon and Carter, 1991; Bouws et al., 1996; Kushnir et al., 1997; WASA Group, 1998).
It appears that recent work points towards increases over time in extra-tropical cyclone activity during the latter half of the 20th century in the Northern Hemisphere, and decreased activity in the Southern Hemisphere. However, the mechanisms involved are not clear, and it is not certain whether the trends are multi-decadal fluctuations, or rather part of a longer-term trend. Furthermore decreased cyclone activity in higher latitudes of the Southern Hemisphere is not obviously consistent with an increase in the positive phase of the Antarctic Oscillation in the last fifteen years or so (Section 2.6.6). A more fundamental question is whether we would expect more or fewer extra-tropical cyclones with increased warming. As pointed out by Simmonds and Keay (2000), the specific humidity increases as temperatures increase, and this increased moisture should enhance extra-tropical cyclones, but Zhang and Wang (1997) suggest that cyclones transport energy more efficiently in a more moist atmosphere, therefore requiring fewer extra-tropical cyclones (see Chapters 7 and 10 for more discussion).
Other reports in this collection