Transcript Slide 1

ESTIMATING
RISK
PROBABILITIES
CENTRE FOR STRATEGIC ECONOMIC STUDIES
BUSINESS AND LAW
Roger Jones
February 26 2009
WWW.VU.EDU.AU
1
Structure of talk
•
•
•
•
The problem with prediction
Estimating risk probabilities
Hedging adaptation and mitigation
How much climate change do we need to adapt to by
when?
2
Scales of approach
Top-down approach
Global
World development
Global greenhouse gases
Global climate models
Regionalisation
Impacts
Vulnerability
(social)
Climate
adaptation
policy
Vulnerability
(physical)
Local
Adaptive capacity
Indicators based on:
Economic resources Technology
Infrastructure Information & skills
Institutions
Equity
Bottom-up approach
Past
Dessai and Hulme 2005
Present
Future
3
Characterising uncertainty in AR4
5
Probability (%)
4
3
2
Likely
1
Very Likely
Virtually Certain
0
-1
0
1
2
3
4
5
6
7
8
9
10
Global Mean Temperature Change (°C)
4
The likelihood of prediction
5
Virtually Certain
Highly Likely
Likely
4
Probability (%)
As Likely As Not
Unlikely
Highly Unlikely
3
Exceptionally unlikely
2
1
0
0
1
2
3
4
5
6
7
8
9
10
Global Mean Temperature Change (°C)
5
The likelihood of risk
.
6
Global mean warming (°C)
5
Remote chance
4
Highly unlikely
Unlikely
3
As likely as not
Likely
Highly likely
2
Virtually certain
1
0
1990
2010
2030
2050
2070
2090
Year
6
Framing climate change risk
Low probability, extreme outcomes
Least likely
Moderately
likely
Considerable
damage to most
systems
Increased
damage to
many systems,
fewer benefits
Highly likely
Almost certain
Damage to the
most sensitive,
many benefits
Happening now
Vulnerable to
current climate
Probability
Consequence
Core benefits of adaptation and mitigation
Probability – the likelihood of reaching or exceeding a given level of global warming
Consequence – the effect of reaching or exceeding a given level of global warming
Risk = Probability × Consequence
7
100
100
75 cm
50 cm
25 cm
80
75 cm
Sea Level Rise (cm)
Sea Level Rise (cm)
80
60
50 cm
40
25 cm
20
75 cm
60
50 cm
40
0
0
0
100
0
Probability (%)
75 cm
60
50 cm
40
25 cm
20
0
80
75 cm
60
50 cm
40
25 cm
20
0
0
5
10
Probability (%)
100
80
75 cm
60
50 cm
40
25 cm
20
0
0
100
Probability (%)
Sea Level Rise (cm)
80
100
Probability (%)
100
Sea Level Rise (cm)
100
Sea Level Rise (cm)
Sea Level Rise (cm)
100
25 cm
20
80
75 cm
60
50 cm
40
25 cm
20
0
0
5
10
Probability (%)
0
100
Probability (%)
8
The new global growth path
• Global growth has accelerated in the past decade, driven by the
developing countries, especially China and India
• This growth is energy and coal intensive, and likely to continue
• Realistic projections of energy use and CO2 emissions to 2030 are
above the SRES marker scenarios, including A1FI
9
Implications for GHG emissions and
atmospheric concentrations
• The implications this new growth path are explored by:
• developing a reference case projections of CO2 emissions from fuel combustion to
2030
• assuming that other emissions grow in a similar manner
• developing policy emissions paths (minimum emissions paths)
• explore CO2-e concentrations and temperatures in a simple climate model
• Minimum emissions paths (MEPs) from 2010 to 2030 were explored in
Sheehan et al. GEC 2007
• The 2030 MEP resembles the SRES A1B “on steroids”
• Current growth to 2100 under reference conditions resembles SRES
A1FI “on steroids”
10
Minimum emissions paths 2010–2030
CO2 emissions (Gt/yr C)
a)
25
20
15
10
5
0
1990
2010
2030
2050
2070
2090
Year
MEP2030
MEP 2025
MEP 2015
MEP 2010
MEP 2020
11
Likelihood of exceedance – range
of reference and policy scenarios
Global mean warming (°C)
.
7
6
5
4
Remote
chance
Highly
unlikely
Unlikely
3
As likely as
not
Likely
Highly likely
2
Virtually
certain
IPCC Low
1
IPCC High
0
1990
2010
2030
2050
2070
2090
Year
12
Framing adaptation
• Goal setting
• Where do we want to go? (aspirational goals)
• How do we want to get there?
• What are the risks?
• What are the barriers? (e.g., lack of adaptive capacity)
13
100
90
80
70
60
50
40
30
20
10
0
en
era
tio
Fo
na
res
le
qu
ts
ity
uc
ce
ss
Br
ion
idg
ed
es
ign
life
Int
erg
An
Elenual
cti cro
on ps
cy
cle
W
s/p
ho
rof
le
it &
Pla farm
los
pla
nt
s
bre
n
nin
To
g
uri edin
Tr sm
g
ee
de cyc
c
les
Ge rop vel
o
ne s
pm
rat
en
ion
Ne
ts
al
w
su
irri
cc
ga
e
tio
n p ssio
Tr
an
n
roj
sp
e
cts
ort
inf
ras
tru
Ma
ctu
jor
re
urb
Pr
ote an i
nf
cte
d a rast
rea ruc
tur
s
e
La
rge
da
ms
Planning horizons
14
90
70
50
40
60
Weak initial response
Strong late adjustment
Strong initial response
Some ongoing adjustment
30
20
10
0
Up-front response
80
Operational pathways
As
pir
ati
on
al
g
oa
l
Operational pathways and
aspirational goals
15
How much climate change needs to
be adapted to by when
Types of climate information required:
• Climate variability (daily to decadal)
• Ongoing rate of change
• Past and near term commitments to climate change
• Climate sensitivity
• Regional climate change projections
• Greenhouse gas emission policies (Mitigation)
16
Reference and policy scenarios for
hedging adaptation and mitigation
Mean Global Warming (°C)
5
Emission scenarios
4
3
2
Climate sensitivity
Past and nearterm commitments
Warming
rate
1
0
1990
2010
2030
2050
2070
2090
Year
Reference
Policy
Observations
17
Hedging adaptation and mitigation –
reference and policy scenarios
Mean Global Warming (°C)
5
Mitigation
benefits
4
MIT-AD
3
AD-MIT
Adaptation
benefits
2
1
0
1990
2010
2030
2050
2070
2090
Year
Reference
Policy
Observations
18
Hedging strategies between
reference and policy scenarios with
high policy uncertainty
Mean Global Warming (°C)
7
6
5
4
3
2
1
0
1990
2010
2030
2050
2070
2090
Year
Adaptation
Ad-Mit
Mit-Ad
Mitigation
A1FI aug–MEP2030
MEP2010–2030
19
Whole of climate approach
• Links current climate and adaptive responses with future possibilities
• Ongoing variability and extremes are the main drivers of current adaptation to
climate, links between variability and longer-term change give these experiences a
future dimension.
• Long-term fluctuations in natural climate variability may be affecting
some regions
• Not all change is anthropogenic
20
Whole of climate approach
An understanding of the dynamics of climate variability is needed to:
• Diagnose fluctuations, shifts or trends as temporary, persistent or
permanent.
• If the dynamics of the change are not understood, statistical or other
methods can be used to explore “what if” questions based on
understandings of climate model and historical behaviour.
21
Regional example of climate changes
– Melbourne, Australia
The Melbourne Region has experienced many step changes rather than
trends
For a 1 by 1 degree area over greater Melbourne:
Annual rainfall :statistically significant downward shift in 1996 in rainfall
from just over 900 mm to 750 mm, -17%.
Max temp: Statistically significant upward step change 1998 of 0.6°C.
About half of this can be explained by the decrease in rainfall (due to a decrease in
cloud cover). About half (0.3°C) is added warming
Analysis of annual frequency of days >35°C and >40°C not significant
All days under 30°C have become significantly warmer
During summer (DJF) almost 1°C warmer
22
Regional example of impacts – southeastern Australia
Streamflow: 60% up to 80% across western Victoria, 25–60% in eastern
Victoria.
Extreme fire weather index (temperature, lower humidity and higher winds):
100 on Black Friday in 1939,
115 on Ash Wednesday in 1983
150–200 on Black Saturday, February 2000
Viticulture: harvest 4–6 weeks earlier, crop losses, smoke damage
Horticulture, dairy: under stress in irrigation regions
Snow: reduced snow cover
Human health (heat stress): hundreds(?) dead from heat stress, 220+ from
fires, event trauma, drought stress in rural regions
Environment: woodland birds decline, tree die-back accelerated, tree planting
failures, icon wetlands critical, frequent hot fires
23
Exploring decision analysis
Post 1997 rain short
term CV + climate
change
Post 1997 rain long
term CV + climate
change
Post 1997 rain –
climate change
p1
p2
p3
Recovery expected
soon, long-term
gradual reduction?
Partial recovery
expected in decades
Serious long-term
deficits
Benefit if correct
Penalty if incorrect
Benefit if correct
Penalty if incorrect
Benefit if correct
Penalty if incorrect
24
Choosing climate information
• Understand risk and risk management options – how is climate
information used in decision-making for specific risks?
• What is my planning horizon and operational pathway? E.g., up front,
incremental, wait and see
• What’s my climate baseline?
• Choose global scenarios based on sensitivity, risk tolerance and
hedging strategies – choose scenarios that are 50% likely to be
exceeded to
• Determine local scaling and down-scaling needs for key climate
variables
• Undertake assessment e.g., modelling, expert analysis
25
Thresholds and key vulnerabilities
• Determine critical limits. E.g., sea level rise, storm severity or surge
protection, flooding, public health limits, water quality and supply
• Diagnose specific climate conditions leading to critical limits
• Establish plausible combinations of change in mean and variability,
natural and anthropogenic, leading to critical thresholds
• Determine likelihood that such conditions may be exceeded within
planning horizons. For cities, many of these horizons will be long-term
26
Caveats and working principles
• All probabilities are subjective – test different plausible assumptions to
test whether outcomes (decisions on risk management) are sensitive to
assumptions
• What information is required to make a specific decision? The less
important climate is compared to other risk factors, the less precision
will be required
• A 1°C warming in 2030 (from 1990) is as likely as not. From 2040+,
considerable hedging between adaptation and mitigation is required.
Without solid emissions policy, hedging for >3°C warming by 2100
needs to be contemplated.
• Sea level rise estimates need to consider outcomes not quantified in
the AR4, including Greenland and perhaps West Antarctica
27
CONTACT DETAILS
ROGER N JONES
BUSINESS AND LAW
CENTRE FOR STRATEGIC ECONOMIC STUDIES
PHONE +61 3 9919 1992
FAX
+61 3 9919 1350
EMAIL [email protected]
WWW.VU.EDU.AU
28