Climate Adaptation: Risk, Uncertainty and Decision-Making

1
Climate AdaptationRisk, Uncertainty and
Decision-Making

• Dr. Robert Willows
• Environmental Forecasting Manager

2
Climate Change
Sequestration
Climate Variability
Emissions
Impacts
Energy Supply
-ve
Adaptation
Energy Demand
Scenarios
Scenarios
Economic and Social Issues and Development
3
Study objectives

• Help identify climate-sensitive decisions

• Help achieve better decisions

• Provides guidance on the use of tools and
  techniques

• Should provide generic guidance on climate risks

• Consistent with DETR guidelines for Environmental
  Risk Assessment (Greenleaves 2)

4
UKCIP Technical Report
Climate Adaptation Risk, Uncertainty and
Decision Making

• Part 1 A review of
• Risk and uncertainty
• Decision-making under uncertainty
• Risk-based climate impact assessment
• Part 2 Framework and Guidance
• Stage by Stage guidance to support the process of
  undertaking risk-based appraisal of climate
  influenced decisions

5
Decision Making Framework
Climate change issue
Climate change policy
Climate change research, monitoring
Identify problem
Establish criteria for decision-making
Monitor
Climate change scenarios
Socio-economic scenarios
Risk assessment
Vulnerability assessment
Data information collection
Implement decision
Impact assessment
Options appraisal
Identify options
No
Make decision
No
Adaptation strategies
Climate change application
Objectives met ?
Climate change policy
Problem defined correctly ?
Yes
Yes
6
Risk Hazard Consequence
7
Hazard Risk and Uncertainty
8
Risk and Uncertainty
9
Risk and Uncertainty
10
Risk Prioritisation Temporally-dynamic risks
11
Climate sensitive decisions
Large
Climate adaptation decisions
Significanceofclimate changeor climate
variable(s)
Climate influenced decisions
Moderate
Climate independent decisions
None
None
Moderate
Large
Significance of non-climate factors or
non-climate variable(s)
12
Decision errors?
13
Adaptation strategies under Uncertainty
Optimistic Precautionary Risk Averse Least
Regret No-regret

• The option the may produce the best
  adaptation outcome

MaxiMax

• The option associated with the most
  favourable of the least favourable
  possible outcomes

MaxiMin

• That option associated with the lowest
  lost opportunities or regret

MiniMaxRegret

• The best adaptation option under all
  possible outcomes

14
Generic options for climate risk management

• Wider use of risk assessment, forecasts and
  options appraisal
• preferably proactive technical response

• Delay and buy-time
• proactive technical response to reduce
  uncertainty

• Research ? e.g. modelling, technology, adaptive
  capacity

• Monitoring
• system performance monitoring - proactive
  technical response
• climate impact monitoring - reactive technical
  response

• Data and information supply, and education,
  awareness raising
• proactive and reactive

• Contingency planning
• low probability, high consequence events
• strategic planning response

15
Generic options for climate risk management

• Diversification or bet-hedging
• proactive technical or policy response

• Insurance - proactive, fiscal response

• Defend and Manage - reactive technical measures

• Change of use
• proactive or reactive, planning response /-
  technical measures

• Retreat and Abandon
• strategic planning response

• Safety factors, climate headroom, buffering
  measures
• technical and regulatory response

16
Adaptation measures and options

• Adaptation not just Defend, Managed Retreat, or
  Abandon
• Informed by formal assessment of climate change
  risk, adaptation options may comprise a portfolio
  of different adaptation measure. These may
  include
• Delay decisions, buy-time, data and
  information gathering, RD, monitoring,
  contingency planning, bet-hedging,
  insurance/other risk spreading strategies,
  safety factors/headroom.

17
Decision Making Framework
Climate change issue
Climate change policy
Climate change research, monitoring
Identify problem
Establish criteria for decision-making
Monitor
Climate change scenarios
Socio-economic scenarios
Risk assessment
Vulnerability assessment
Data information collection
Implement decision
Impact assessment
Options appraisal
Identify options
No
Make decision
No
Adaptation strategies
Climate change application
Objectives met ?
Climate change policy
Problem defined correctly ?
Yes
Yes
18
See associated notes pages...
19
Risk screening - climate variable checklist

• Helps to both identify (Table 1) and define the
  different characteristics (Table 2) of
  potentially significant or relevant climate
  variables
• Includes preliminary assessment of sensitivity
  and confidence
• Useful for screening of variables
• Not constrained by availability of climate
  forecast variables (e.g. from GCMs or RCMs)
• Encourages rigorous analysis of climate influence
• Table 1 is not complete - proxy and compound
  variables will depend on nature of particular
  assessment

20
Climate variable checklist
Types of variables

• Examples
• Primary CO2, sea-level, temperature,
  precipitation, wind, cloud cover
• Synoptic Weather types, pressure, storm track,
  lightning
• Compound Humidity, evapotranspiration, mist,
  fog, growth season
• Proxy Soil Moisture, river flow, wave climate

21
Climate variable checklist
Characteristics of variables

• Examples
• Magnitude and Direction Increase, decrease, rate
  of change
• Statistic Average, time-integrated,
  variability and frequency
• Averaging period Instantaneous ... hourly .
  Annual ..decadal
• Joint probability events Consecutive,
  coincident or joint occurrence,
• and variables correlation

22
(Climate) Influence diagrams
23
Describing confidence ..
24
Downscaling 1

• Space
• GCM ? site
• Time
• Monthly ? daily

25
SDSM - Statistical Down-Scaling Model (Rob
Wilby, Kings College London)

• Daily data - observed data at site
• Model - site data and large-scale data from GCM
• Scenario - generate ensembles of daily time
  series

26
Downscaling - Flood return period prediction
27
Scenario analysis and risk assessment
Q. Can we create scenarios which reflect changes
in variability as well as the mean?
A. YES - but it is difficult and scenarios
remain contingent on assumptions and
non-quantified uncertainties
Q. Can we assign probabilities to different
scenarios?
A. Probably. Expert judgement can be used to
assign probability to the range encompassed by
any two scenarios but uncertainty components
within suite of scenarios have to be well-posed
28
Estimating probabilities of different futures
We typically use a small number of scenarios
If we make assumptions about the likelihood of
different emissions futures, use many more
climate models, and incorporate the effects of
natural climatic variability, we can generate
many more scenarios
29
Probabilistic scenarios
1
(Subjective) Probability
0
L-gtM
L-gtMH
L-gtH
Scenario
30
Incorporating climate change into water resources
management
The future hydrological resource base will not
be the same as the present resource base
Mean climate will be different, due to climate
change and natural climatic variability
Variability in climate will be different.
Altered frequency of successive dry years?
.but we dont know how different.
31
Influence diagrams - water resource management
management objectives
demand
land use change
Supply-side
water resources model
hydrological model
baseline data
reliability
Demand-side
Climate change
adaptive response
32
Use of socio-economic scenarios
Water resources for the future
A STRATEGY FOR ENGLAND AND WALES
March 2001
33
Coping with uncertainty
1. Flexible management approaches - review
situation and adjust plans if appropriate -
continued monitoring
2. Improved seasonal forecasts - based on
understanding of causes of seasonal
climatic variability
3. Scenario analysis and risk assessment
34
Conclusions and recommendations (for
decision-makers)

• Emphasis on understanding impact of present-day
  observed climate variability

• Future climate change is only one source of
  decision uncertainty

• Assessment of climate risk should be hierarchical
  /tiered

• Climate adaptation should be iterative

• Assumptions and sources of uncertainty should be
  treated explicitly in risk and impact assessments
  in order to reach robust decisions