Human response to climate change - The Global Climate System: Patterns, Processes, and Teleconnections - page 270

and much existing loss information is based on estimates rather than actual
measurements.
This essay focuses on results that have carefully assessed the economic
impacts from climate conditions during the past 50 years. Fortunately, recent
studies of past impacts have made careful adjustments to the raw historical
data. Studies of past economic impacts require careful attention and adjust-
ments for shifting changes in the ''target'' such as varying crop varieties/
hybrids over time, for inflation, and for other changing societal and techno-
logical conditions that affect the measurement of impacts. For example,
comparison of loss data from a drought in 1950 and losses from a similar
drought in 1988 requires adjustment for the changes in the area's population
density, in the types of water resource systems, and the levels of insurance
coverage.
This essay has five sections. First, the temporal behavior of various
extreme climate conditions during 1950-2000 is assessed. Next, national
annual losses from damaging extremes, including floods and hurricanes,
during 1949-97 are reviewed. In the third section, financial losses and
gains are identified for recent major climate extremes, including the 1988
drought, the 1993 flood, and the extremely warm-dry winter from El Ni˜o
1997-8. Factors causing the increased losses during recent decades appear
in part four. The final section summarizes national losses and gains, how the
impacts rate in the nation's economy, and potential future impacts resulting
from possible future climate changes.
8.4.2 Temporal fluctuations in climate extremes
Individual climate conditions have been assessed in a variety of studies. This
information provides a basis for comparing the temporal behavior of certain
conditions and the impacts created. If losses related to a given climate con-
dition are increasing but the condition itself is decreasing, this difference
helps point to other non-atmospheric factors affecting the amount of loss.
Heavy precipitation and flooding
Recent studies have defined trends to more frequent heavy precipitation
events. Karl et al.( 1995 ) found that 1-day heavy precipitation events exceed-
ing 5.1 cm have made an increasingly large contribution to annual precipita-
tion over the United States since 1910. Heavy precipitation events of 7-day
duration are closely related to hydrologic flooding occurrences on small to
medium-sized rivers, and trends in 7-day heavy precipitation events for the
entire country have increased in recent decades (Kunkel et al. 1999 ).
Lins and Slack ( 1997 ) reported upward streamflow trends, consistent with
the observed upward trends in heavy precipitation. Extensive human