Human response to climate change - The Global Climate System: Patterns, Processes, and Teleconnections - page 271

modification of river basins makes it very difficult to assess long-term
trends in flood flows on a national basis and relate this to flood damage
(Downton and Pielke 2001 ). The complexities of shifting climate condi-
tions, altered drainage systems and river courses, plus changes in societal
vulnerability make identification of the role of changing climate conditions
difficult to quantify as the primary cause of changes in flood impacts.
Available data indicate that flood-related damage has increased in recent
decades.
Hurricanes
The frequency of land-falling intense hurricanes has decreased since 1950
(Kunkel et al. 1999 ). Figure 8.8 shows that an increase in hurricane damage
(unadjusted for inflation and land-use changes) over recent decades has
occurred during this period of decreasing hurricane frequencies and intensi-
ties. This means that fewer storms are responsible for the increased damage,
and these storms are no stronger than those of past years. Adjustment of losses
for temporal changes in population growth and development in vulnerable
coastal locations revealed that losses have not increased over time, and these
factors are the key to a steady series of losses, rather than increased storm
numbers and strength (Pielke and Landsea 1998 ).
Thunderstorms, hail, and tornadoes
The national average number of thunderstorm days during 1910-2000 has a
downward trend, decreasing from 40 days in 1910 to 38 days in 2000. The
thunderstorm decline agrees with the downward trend in national thunder-
storm losses over time (Changnon and Hewings 2001 ). Changnon ( 2001 )
found that thunderstorm-created losses had increased in Florida and the West
Coast, but storm frequencies had not increased in these areas. This suggested
that the increased losses were a result of the large regional growth of popula-
tion and wealth in these areas.
The average number of hail days per year during 1910-2000 also has a
downward trend. This agrees with a downward trend in the crop-hail losses
(Changnon et al. 2001 ). Property losses from hail had a major increase during
the past 15-20 years, mainly in rapidly expanding cities. The number of
tornado days nationally during 1953-97 had a slight increase over time until
1970, and a flat trend thereafter. Tornado losses showed a temporal increase,
but the national number of killer tornadoes, those tornadic storms leading
to one or more deaths, during 1953-97 show a marked temporal decrease
(Changnon and Hewings 2001 ). The increases noted for all tornadoes and
their losses are considered a result of growing population and more attention
to tornado occurrences, collectively leading to more tornadoes seen over time
(Kunkel et al. 1999 ).