Changnon (1996a) studied the effects of potential shifts in summer precipitation
on transportation in Chicago, using data for 1977–79 and assuming continued
use of current modes of transport. The study suggests that a future climate
with more summer rainy days, somewhat higher rain rates, and more rainstorms
would increase total vehicular accidents and total injuries in vehicular accidents,
reduce travel on public transportation systems, and cause more aircraft accidents
and delays. A drier climate probably would experience fewer moderate to heavy
rain events, but results show that rain events during drier conditions produce
a greater frequency of accidents and injuries per event than during wetter conditions.
If high-heat events became more common with warmer climate, they also could
become a problem. They have been known to soften asphalt roads, “explode” or
buckle concrete roads, warp railroad rails, close airports because of lack of
“lift” in extremely hot air, and increase mechanical failures in automobiles
and trucks. On the other hand, there might be fewer mechanical failures resulting
from extreme cold (Adams, 1997). Floods are costly to transportation systems,
as they are to other infrastructure. Although the effect of climate change on
flying weather is not clear, transportation by air is known to be sensitive
to adverse weather conditions; major systemwide effects sometimes follow from
flight cancellations, rerouting, or rescheduling. For example, one diverted
flight can cause anywhere from 2 to 50 flight delays, and one canceled flight
can result in 15–20 flight delays. The cost of a diverted flight can be as much
as US$150,000, and a cancellation can cost close to US$40,000. The corresponding
direct annual costs to 16 U.S. airlines are US$47 million and US$222 million,
respectively (Qualley, 1997). Several additional examples of impacts on transportation
are cited in Chapter 13.
Flooding and other extreme weather events that damage buildings and infrastructure
could cost the world’s economies billions of dollars under climate change simply
to replace the damage—a cost that could divert funds from other needed investment
(see Chapter 8). However, Mimura et al. (1998) note that
cost increases for disaster rehabilitation and countermeasures against natural
calamities could expand the market for the construction industry. Although no
direct studies have been done, it is likely that a greater incidence of summer
heat waves would reduce the productivity of this sector, but a lower incidence
of cold waves and snowy conditions would increase the amount of year-round construction
that could be accomplished in climates that currently have long, cold winters.
Changes in design requirements for infrastructure, leading to additional requirements
for construction, are discussed in the SAR.
Manufacturing industries that are not directly dependent on natural resources
generally would not be affected by climate unless key infrastructure is destroyed
by flood or landslides, or unless shipments of inputs and outputs are affected
(e.g., by snow blocking roads, airports, and train tracks; flooding or low flow
that make river transportation untenable; or low water supplies that make process
cooling and environmental activities more difficult). However, manufacturers
are influenced by climate change in two other ways. First, they would be affected
through the impact of government policies pertaining to climate change, such
as carbon taxes (thereby increasing the cost of inputs). Second, they could
be affected through consumer behavior that in turn is affected by climatic variations.
For example, less cold-weather clothing and more warm-weather clothing might
be ordered. Manufacture that depends on climate-sensitive natural resources
would be affected by impacts on those resources. For example, food processing
activity would follow the success of agriculture. Very little is known concerning
the effects of warming on industry, and most information is highly speculative.
7.3.5. Financial Services and Insurance
Climate change increases risks for the insurance sector, but the effect on
profitability is not likely to be severe because insurance companies are capable
of shifting changed risks to the insured, provided that they are “properly and
timely informed” on the consequences of climate change (Tol, 1998). For example,
during the great storms in the early 1990s, the insurance sector reacted to
increased risk and large losses by restricting coverage and raising premiums.
Tucker (1997) also shows that increased climatic variability necessitates higher
insurance premiums to account for the higher probability of damages. However,
insurance companies still can be destabilized by large losses in a major weather-related
catastrophe in a region where actuarial tables and estimated risks do not adequately
reflect true weather risk (including greater variability), and companies therefore
may not have made adequate provision for losses. See Chapter
8 for a description of impacts on financial services.
7.3.6. Estimating and Valuing Effects
Valuation of climate impacts remains difficult on three grounds. The first
is uncertainty associated with determining physical changes and responses to
these changes. The second is economic valuations of these changes that vary
across regions. Fankhauser et al. (1998) show that damage cost estimates are
sensitive to assumptions made on the basis of valuation (willingness to pay
versus willingness to accept), accountability for impacts, differentiation of
per unit values, and aggregation of damage costs over diverse regions. A third
problem can be expressed as follows: “Which metric?” Five popular metrics are
used: market costs, lives lost, species lost, changes in the distribution of
costs/benefits, changes in quality of life (loss of heritage sites, environmental
refugees, etc). Schneider et al. (2000) conclude that when aggregation exercises
are undertaken, disaggregation of all estimated effects into each of five numeraires
is needed first, followed by a traceable account of any aggregation so others
holding different weighting schemes for each numeraire can re-aggregate. This
is done rarely, if ever.
The Workshop on the Social and Economic Impacts of Weather at the National
Center for Atmospheric Research, 2–4 April 1997, in Boulder, CO, estimated that
property losses from extreme weather of all types currently costs the United
States about $15 billion yr-1 ($6.2 billion related to hurricanes),
as well as about 1,500 deaths (about half resulting from cold events); the worst
flood and hurricane years yield about $30–40 billion in property losses.
Smith (1996) standardized estimates of climate change damages for the United
States for a 2.5°C warming, a 50-cm sea-level rise, 1990 income and population,
and a 4% real rate of return on investments. Total damage estimates are slightly
less than 1% of United States gross national product (GNP) in 1990. Within individual
sectors such as agriculture and electricity, however, standardized damages differ
by more than an order of magnitude. This level of uncertainty appears to apply
among experts as well. For example, Nordhaus (1994) surveyed experts, and their
damage estimates ranged over more than an order of magnitude.
Yohe et al. (1996) calculated the cost of a 50-cm sea-level rise trajectory
for developed property along the U.S. coastline. Transient costs in 2065 were
estimated to be approximately $70 million (undiscounted and measured in constant
1990 US$). These costs are nearly an order of magnitude lower than estimates
published prior to 1995 (e.g., Fankhauser, 1995). This is because Yohe et al.
(1996) incorporated the cost-reducing potential of market-based adaptation in
anticipation of the threat of sea-level rise. In addition, they assumed efficient
discrete decisions to protect or abandon small tracts of property, based on
their economic merit. Some work since suggests that maladaptation may cause
the costs of sea-level rise to be somewhat higher (West and Dowlatabadi, 1998).
7.3.7. Tools/Methods/Approaches/Models Used in Developing
New Knowledge, including Assumptions, Sensitivities, and Scenarios Used in Models
Current impact assessment methods focus on comparing current conditions to
a single alternative steady state—that associated with doubling of GHGs. Mendelsohn
and Schlesinger (1999) attempt to estimate climate response functions for market
sectors in the United States that reflect how damages change as climate changes
through a range of values. Impacts are generated by using national climate values,
rather than global values, and the timing of climate change is included in the
modeling of capital-intensive sectors such as coastal resources and timber,
which cannot adjust quickly. Empirical estimates of climate response functions
are based on laboratory experiments coupled with process-based simulation models
and cross-sectional studies (Mendelsohn and Neumann, 1998). Both methods indicate
that agriculture, forestry, and energy have a bell-shaped relationship to temperature.
Similarly, an increase in precipitation is likely to be beneficial to some agriculture,
forestry, and water sectors, although this effect is reversed at sufficiently
high levels. However, this work captures neither the transient response of the
climate system nor the actual dynamics of the energy sector in response to climate
(e.g., Schneider, 1997).