Transcript Estimating the Global Damages from Climate Change

JOEL SMITH
VICE PRESIDENT
STRATUS
CONSULTING
INC.
Estimating the Global Damages
from Climate Change
Joel Smith
and
Sam Hitz
Stratus Consulting Inc.
October 23, 2002
Objective
• Effort to identify benefits from different greenhouse
gas emissions policies
 key issue is identification of marginal benefits
(avoided damages); are they:
 constant
 decreasing
 reverse sign?
 provide useful input for examining long run goals
• Did not address the costs of mitigation
Overview
• We surveyed published sectoral studies that quantify
the global impacts of climate change.
 Limited ourselves to global studies; did not examine regional
impacts literature
• Characterized the relationships between climate
change and impacts based on studies results
 increasing impacts
 parabolic
 indeterminate
• Identified key factors and assumptions in these
studies that could substantially affect results
 inclusion and correctness
Methodology and Approach
• Global mean temperature (GMT) used as indicator of
climate change
 recognize the many caveats with doing so
• Use equilibrium (generally older studies) and transient
(generally newer studies) results
• Used metrics employed by authors of individual
studies. No attempt to aggregate.
• Identified and analyzed key factors, assumptions and
framework upon which studies were based.
Sectors Impacted by Climate
Change
•
•
•
•
•
•
•
•
•
Published Studies
Agriculture
Sea Level Rise
Water Resources
Human Health
Terrestrial Ecosystems
Forestry
Marine Ecosystems
Biodiversity
Energy
•
•
•
•
•
•
No Global Studies
Recreation and Tourism
Transport
Building
Insurance
Human Amenity
Mitigation
Presentation
• Results from agriculture, coastal resources,
biodiversity and water sector are representative of
these categories.
Agriculture
Rosenzweig et al., 1995
Percent Change in Number of People at Risk of Hunger (2060)
50
641 million people at risk in 2060
Reference Scenario
% change in number of people
40
UKMO
Adaptation Level 1:
Shifts in planting date (+/- 1 month)
Additional application of irrigation water to crops already under irrigation
Changes in crop variety to currently available varieties
30
Adaptation Level 2:
Large shifts in planting date (>1month)
Increased fertilizer application
Installation of irrigation systems
Development of new crop varieties
20
UKMO
GFDL
10
Level 2 Farm Adaptation
Level 1 Farm Adaptation
Low Temp. No Farm Adaptation
0
0
1
GISS
GFDL
2
3
-10
GISS-A
2030
-20
°C
4 GISS
5
6
Agriculture
Parry et al., 1999
Percent Change in Number of People at Risk of Hunger
60
2080s
642 ppmv CO2
HadCM2
HadCM3
Percent change in number of people at risk
50
250 million people at risk of
hunger in reference
scenario.
40
2080s
731 ppmv CO2
30
2050s
527 ppmv CO2
2020s
443 ppmv CO2
20
2020s
441 ppmv CO2
10
2050s
565 ppmv CO2
0
0
0.5
1
1.5
2
°C
2.5
3
3.5
Agriculture Conclusions
• Results suggest initial benefits that eventually
decrease and give way to damages as GMT rises, or
alternatively, initial damages that decrease before
continually rising.
• Some disagreements at lower temperatures, but
eventually increasing damages beyond 3-4° C.
• This result agrees with expectation based on
underlying biophysical processes.
Coastal Resources
Fankhauser, 1995.
Cost of Sea Level Rise in OECD Countries
2000
Billions $
1500
1000
500
0
20
100
Sea level rise by 2100 (cm )
200
Coastal Resources
Nicholls et al., 1999
Additional People in the Hazard Zone as a Function of SLR
70
HadCM2
HadCM3
2080s
60
Additional people (millions)
50
40
2050s
30
20
2020s
10
0
0
5
10
15
20
25
Sea level rise (cm )
30
35
40
45
Coastal Resources Conclusions
• Damages seem to increase linearly with SLR.
• This result is in line with expectations of monotonically
rising costs.
• These costs likely to continue accruing well into the
next century, as sea level continues to rise, even after
CO2 stabilizes.
Biodiversity
Halpin, 1997.
Percent Change in Eco-Climatic Classes for Biosphere Reserves
Compared to Global Average
90
Biosphere Reserves
Total Terrestrial Area
% of area where eco-climatic class change occurs
80
70
UKMO
60
GFDL
GISS
50
OSU
40
30
20
10
0
0
1
2
3
°C
4
5
6
Biodiversity
• Eco-climatic classes change within global bioreserves with greater frequency as GMT rises.
• There is little reason to doubt this result. It would be
difficult to argue that climate change will slow the loss
of threatened species.
• This is mentioned because even loss of individual
species or ecosystems could be motivation for
mitigation.
Water Resources
Arnell, 1999.
Change in Number of People in Countries in Water Stress
150
HadCM2/GGax
HadCM3/GGa1
100
Number of people (millions)
2020s
2080s
50
2050s
0
0
0.5
1
1.5
2
-50
-100
°C
2.5
3
3.5
Water Resources
Arnell, 1999.
Difference Between Total Population in Countries Where Water Stress
Increases and Countries Where Water Stress Decreases
1500
HadCM2/GGax
HadCM3/GGa1
2080s
1000
Number of people (millions)
2050s
2020s
500
0
0
0.5
1
1.5
2
-500
-1000
-1500
-2000
-2500
-3000
°C
2.5
3
3.5
Water Resources Conclusions
• No clear relationship between climate change and
impacts on water resources.
• Averaging at the regional or country level presents
problems. Basin level is more appropriate.
• We think there should be increasing damages with
increasing GMT.
 Not (yet) borne out by published literature.
Aggregate Studies
3
2
1
Percent of World GDP)))
0
-1
-2
-3
-4
-5
-6
Mendelsohn, output
-7
Nordhaus, output
-8
Nordhaus, population
-9
Tol, output
-10
Tol, equity
-11
0
1
2
3
4
5
Global Mean Temperature (°C)
6
Sectoral Damage Relationships
Table 6. Summary of sectoral damage relationships with increasing temperature.
Sector
Increasing damagesa
Agriculture
Coastal
Parabolic
Unknown
X
X
Water
X
Health
X
Terrestrial ecosystem productivity
X
X?b
Forestry
Marine ecosystems
Biodiversity
X?c
X
Energy
X
Aggregate
X
a. Increasing damages means there are damages with small increases in GMT, and the
damages increase with higher GMTs. We are unable to determine whether the damages
increase linearly or exponentially with GMT.
b. We believe this is parabolic, but with only one study it is difficult to ascertain temperature
relationship, so there is uncertainty about this relationship.
c. This relationship is uncertain because there is only one study on this topic.
Limitations
• Did not consider:
 impacts of temperature changes in excess of 1.4 to 5.8° C
range considered likely by Houghton et al.,2001
 change in climate variability
 impacts due to long term (i.e. post 2100) climate change
 potential large scale singular events (e.g. collapse of THC or
WAIS)
 interaction among impacts on different sectors (e.g. water
and agriculture)
 ancillary benefits and proactive adaptation.
Limitations
• Also:
 Adaptation generally handled with simplistic
assumptions
 can over or underestimate impacts
 Assumptions about population and development
not consistent
 development can make substantial difference in
vulnerability of societal sectors
Conclusions
• In all sectors, the results point to increasing damages
beyond a temperature range of 3-4°C.
• Below this critical temperature range, the picture is
much less clear. In some sectors damages may
accrue immediately and continue to grow. In others,
benefits may eventually give way to damages or initial
damages may decrease before steadily increasing.
• Aggregate studies we examined, tend to confirm this
basic conclusion.
Final Caveats
• Changes in the key assumptions and simplifications
upon which each of the studies depend could either
lower or raise this critical temperature range.
(e.g. adaptation, development, treatment of variability,
interaction among sectors, long-term temperature change
or changes in the climate system)
• We should expect significant variation among regional
results. The critical temperature range we identify
could be quite different depending on the particular
combination of region and sector.