Download Climate Change and Low Income Countries Channing Arndt

Survey
yes no Was this document useful for you?
   Thank you for your participation!

* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project

Document related concepts
no text concepts found
Transcript
Climate Change and
Low Income Countries
Channing Arndt
United Nations University –
World Institute for Development Economics Research
and
University of Copenhagen
With contributions from many others
Risk & Uncertainty
• Knight (1921) :
– “risk" refers to situations where the decision-makers can
assign mathematical probabilities to the randomness which
they face.
– "uncertainty" refers to situations when this randomness
"cannot" be expressed in terms of specific mathematical
probabilities.
MIT JP – Uncertainty Approach
• Integrated Global Systems
Model –IGSM
• 70 Uncertain parameters
• 400 Monte Carlo
• Climate & Socio-Econ
MIT JP – Uncertainty to Risk
Webster et al. (2010). MIT Joint Program Report #180.
Observations
• Projected temperature rise to 2100 is much more
severe than it was only a few years ago due to:
– Actual emissions growth.
– Sensitivity of the climate system.
• Very high levels of warming are possible.
• Mitigation policy is powerful.
– Expected temperature outcomes lower
– Extreme temperature outcomes eliminated.
• Considerable warming appears to be unavoidable.
Adaptation policy also required.
Evolving View of Climate Change
1. Environmental issue
2. Humanist moral issue
3. Global security issue
UNU-WIDER views climate change as a global security issue.
Adaptation to Climate Change
Precipitation 2100: CCSM v. MIROC
Some Regularities
• Warmer – Higher temperatures will be observed
throughout the globe.
• Wetter – Due to speeding of the hydrologic cycle,
precipitation globally is likely to increase. The distribution
of this increase is unknown.
– Note an indeterminate affect on the climate moisture index
[CMI =f(P/T)].
• More intense. More intense rainfall and higher
probability of extreme weather events (flooding, perhaps
cyclones…).
Recent African country studies
• Malawi (2010)
– Analysis of extreme weather events (i.e., droughts and floods)
– Adaptation via drought-resistant seed varieties
• Zambia (2009):
– Contrasts historical variability and 3 climate change scenarios
– Focuses exclusively on agriculture
• Tanzania (2010):
– Focus on agriculture and income distribution
– Various crop modeling approaches
Economics of Adaptation to Climate Change
• Mozambique, Ghana, and Ethiopia (2010):
– Considers 4 future climate change scenarios
– Multi-sector approach: agriculture, energy, infrastructure, etc
– Initial focus on the costing of adaptation options for COP15 in
Copenhagen.
• Current expansion into detailed country studies.
Focus on Mozambique
1. Analytical Framework
Climate change projections
Temperature
Precipitation
PET
Hydrology
Stream-flow
Runoff
Net Evaporation
Floods and Inundation
Flood recurrence
Crops
Hydropower
Infrastructure
Crop yields
Production levels
E.g., road etwork length
Dynamic Economy-wide CGE Model (WIDER)
National/sector GDP
Regional production
Household welfare
2. Climate change impacts
Four climate scenarios
• Global wettest and driest scenarios
• Local wettest and driest (i.e., for Mozambique)
• Selected using “Climate Moisture Index” (CMI)
– Global wet/dry are actually dry/wet scenarios for Mozambique
– Global dry scenario is very wet scenario for Zambezi River Basin
and SADC
• Must use multiple GCMs
• Must take regional approach
Scenario
CMI
Global wet
-0.6
Global dry
+9.3
Local wet
+33.0
Local dry
-58.6
2. Climate change impacts
Agriculture and crop yields
• CLICROP models: 14 crops in 3 sub-national regions
• Captures daily T and P effects, water-logging and
irrigation water demand (exclude CO2 fertilization)
Average change in yield from baseline, 2041-2050 (%)
Global Wet Global Dry
Cassava
North region
Center region
South region
Maize
North region
Center region
South region
Local Wet
Local Dry
2.01
-4.75
-9.36
-3.44
-6.24
-3.27
-0.09
-3.10
0.36
-6.51
-6.21
-3.20
1.27
0.34
3.49
-1.32
0.64
6.37
-2.92
-3.04
-4.36
-1.87
-5.59
-3.95
2. Climate change impacts
Hydropower generation
• IMPEND model determines hydropower generation
based on streamflow and installed/planned investments
• Hydropower declines in all scenarios, but Mozambique
remains a net energy exporter regardless
Average annual production (Giga watt hours per year)
2003-2010
2011-2020
2021-2030
2031-2040
2041-2050
Baseline
Global Wet
13,533
1.09
17,391
-2.35
26,991
-1.82
26,087
-3.94
25,479
-3.37
Change from baseline (%)
Global Dry
Local Wet
0.26
-3.07
-0.55
-7.36
0.40
-5.30
-0.62
-8.08
-0.98
-4.15
Local Dry
-5.31
-6.62
-6.75
-7.26
-12.04
2. Climate change impacts
Flooding and road infrastructure
• CLIRUN: River basin models predict change in flood RPs
• CLIROAD: Captures P, T and flood damages on roads
• Global Dry has most flooding (regional basin effect)
• This damages roads more than in other scenarios
Global Wet
Global Dry
Local Wet
Local Dry
Change in national road
network length relative
to baseline, 2050 (%)
-16.1
-22.4
-11.9
-2.1
Beira SLOSH Model Setup
• Wind
fields from the
storm generation
step generate storm
surges in SLOSH
• This “snapshot”
shows storm surge
above sea level at a
time when a storm is
offshore
Effect of SLR on Return Period for the 100-year
Storm Surge in Beira
Return Time for Current 100 Year Flood
120
100
80
y = 105.65e-3.272x
R2 = 0.9948
60
40
20
0
0
0.05
0.1
0.15
0.2
Low
SLR (meters)
0.25
0.3
Medium
0.35
0.4
High
2. Climate change impacts
Dynamic CGE model
Tanzania
Detailed economic structure:
•4 regions (3x rural, urban)
•33 sectors (17 in agriculture)
•7 factors (3x land, 3x labor, capital)
•20 households (rural/urban quintiles)
Malawi
Zambia
Zimbabwe
Recursive dynamic:
•Capital accumulation on past investment
•Exogenous TFP (linked to sector models)
•Autonomous adaptation (“typical farmer”)
2. Climate change impacts
Baseline “no climate change” scenario
• Baseline specifies a future scenario reflecting development
trends, policies, and priorities without climate change.
• Consistent with sector
models’ baselines (i.e.,
CGE captures individual
baselines and their
interactions within a
market economy)
Share of total GDP (%)
• Assumes a reasonable trajectory for growth and structural
change until 2050 (e.g., falling agricultural GDP share).
35
30
Agriculture
Industry
25
20
15
3.7% average annual GDP growth
2. Climate change impacts
Economywide damages
Effects of climate change are negative and grow with time
Large declines in national welfare by 2050
Wide variation in impacts across CC scenarios
In worst scenario adaptation cost is about US$7.6 billion ($410 mil.
p.a.)
Cumulative deviation in total absorption, 2003-2050
(5% discount rate)
8
Discounted US$ billion (constant 2003
prices)
•
•
•
•
7
6
2.5
5
4
2040s
2.1
1.8
1.4
2020s
3
2
1
0
2030s
1.6
1.1
1.5
1.1
0.6
0.7
0.6
0.5
0.4
Global dry (CSIRO) Global wet (NCAR) Moz dry (UKMO)
1.2
2010s
1.0
2000s
0.7
Moz wet (IPSL)
2. Climate change impacts
Decomposition of impact channels
• Road network is the main impact channel due to major flooding
within the trans-boundary water basin.
• Crop yield damages are most severe in Local Dry scenario.
• Hydropower reduces surplus energy exports but not welfare.
Cumulative deviation in total absorption, 2003-2050
(5% discount rate)
Change in per capita absoprtion growth
rate from baseline (%-point)
Global dry (CSIRO)
Global wet (NCAR)
Moz dry (UKMO)
0.0
-0.1
-0.2
-0.3
-0.4
Falling crop yields and rising sea level
Detriorating transport system
Declining hydropower generation
Moz wet (IPSL)
3. Adaptation investments
Step 1: Transport system investments
• Sealing unpaved roads reduces worst case CC damages by 1/5 with little or no
additional costs (i.e., advisable even under baseline).
Step 2: Irrigation investments
• 1 million hectares of new irrig. Land only slightly reduces CC damages.
Step 3: Agricultural R&D or education investments
• Raising agricultural productivity by a bit more than 1% each year offsets
remaining damages.
• Increasing rate of human capital accumulation also offsets damages.
Baseline
No climate
change
Global dry
Global wet
Moz dry
Moz wet
(1)
2.11
2.11
2.11
2.11
Impact
With
climate
change
(2)
1.73
1.85
2.02
1.91
Adaptation scenarios
Transport
infrastructure
(3) = (2+)
1.81
1.92
2.04
1.97
Agriculture
R&E
Irrigation
expansion
Education
(4) = (3+)
2.11
2.22
2.32
2.27
(5) = (3+)
1.84
1.95
2.07
2.00
(6) = (3+)
2.11
2.23
2.35
2.28
6. Summary of results
• Without public policy changes, the worst scenario results in a
net present value of damages of about US$7.6 billion.
– Equal to an annual payment of US$420 million (5% discount rate).
• Hardening rural roads reduces worst case impacts
substantially, restoring approximately 1/5 of lost absorption.
• Remaining welfare losses could be regained with improved
agricultural productivity or human capital accumulation.
• Investments costs required to restore welfare losses are
subject to debate, but are reasonably less than US$400
million per year over 40 years.
6. Summary of Results
• It is unlikely to be cost effective to protect the vast
majority of coastal regions of Mozambique.
• High value and vulnerable locations, such as the port of
Beira, merit specific consideration.
• Even relatively small levels of sea level rise dramatically
increase the probability of severe storm surge events
(assuming no change in the intensity and frequency of
cyclone events).
6. Policy Recommendations
• Best adaptation to CC is more rapid development leading
to a more flexible and resilient society.
– Effective strategies should reinforce development objectives
• But climate-specific interventions include:
–
–
–
–
Regional adaptation strategies (e.g., river basin management)
Agricultural research & extension
Seal unpaved roads (makes sense even if no CC)
Soft adaptation where possible (e.g., Land use planning: most
capital in low-income countries has not yet been invested)
– Hard adaptation should be heavily scrutinized (e.g., dykes may
reduce risk but increase exposure)