Transcript IASA_Flood & drought_Stefan Hochrainer

Vulnerability and disaster risk mapping workshop
EEA, Copenhagen, 2 July 2009
Assessing and Managing Europe's
current and future flood and drought risks
Hochrainer Stefan
International Institute for Applied System Analysis
Laxenburg, Austria
Point of Departure
Number of events
100
Drought
90
Earthquake
80
Extreme Temperature
70
Flood
60
Volcano
Wild Fires
50
Wind Storm
40
Total
30
20
10
19
70
19
72
19
74
19
76
19
78
19
80
19
82
19
84
19
86
19
88
19
90
19
92
19
94
19
96
19
98
20
00
20
02
20
04
20
06
0
Years
Large Scale Events:
*
Odra Flood, 1997: 5.0 billion Euro losses,
0.8 billion Euro insured losses
300.000 people evacuated
Flooding, 2002: 14.4 billion Euro losses,
3.4 billion Euro insured losses
400.000 people evacuated
* Source: CRED, 2008
Point of Departure
• Losses from weather extremes such as floods, droughts, and other
climate-related events in Europe (and elsewhere) have escalated
in recent decades
• The increase has been more rapid than population or economic
growth can fully account for
• According to the IPCC Fourth Assessment Report, anthropogenic
climate change is expected to lead to increases in intensity and
frequency of weather extremes
Point of Departure
•
Europe vulnerable to disasters already today, key focus for EU adaptation
strategy (White paper) is on managing climate variability
•
Large events:
– 2002 large-scale flooding over central Europe: losses > 15 billion Euro
– 2003 summer heat wave of unprecedented magnitude resulting in
80,000 additional deaths
– led to agricultural losses exceeding 13 € billion and a 30 per cent
reduction in gross primary production of terrestrial ecosystems.
•
Risk information (maps) exist in some EU member states, but ADAM
provides the first comprehensive probabilistic maps of riverine flood and
drought/heatwave risks across the EU on various scales, e.g. from the
regional to the national level
•
Potential applications
– Risk based planning: e.g., flood protection
– Identification and comparison of vulnerable and “at risk” countries
– Financial and economic planning: Risk sharing
Methodological Approach
• Extremes are low probability -high consequence events.
• To assess and manage extreme risks probabilistic approaches
have to be used to incorporate all possible future scenarios.
• Traditional risk measures are inadequate for decision making, e.g.
averages will not give adequate representation of the risk.
• The „fat-tails“ and the thickness (of distributions) are important instead.
• Traditional coping mechanisms do not work in the case of
extremes, e.g. Law of large numbers not applicable; hence
different risk management/adaptation strategies have
to be considered for catastrophes.
• Distinctions between stock and flow effects are important.
Advantages of Risk based approaches
Efficiency of risk management instruments dependent on the
occurrence probability
Low probability
EU solidarity fund
Probability
Market based insurance
Flood protection
High probability
Flood
Risk based approach: Assessing direct risk
Risk is a function of the
Hazard, the Exposure and
the Vulnerability.
Approach used in ADAM:
Risk based approach: Assessing direct flood risk
Maximum annual average
flood damage for European
provinces and regions
(NUTS 2 level) as a
percentage of GDP for
today’s climate regime
Averages are based on
loss distributions on the
GRID level.
New method developed to
upscale losses to the national
level, „hybrid-convolution“
Method (Hochrainer, Lugeri)
Assessing direct flood risk: Results
2
Minimum Average Annual Damage in %GDP
1.8
Maximum Average Annual Damage in %GDP
Percent of loss to GDP
1.6
1.4
1.2
1
0.8
0.6
0.4
0.2
0
UK
SK
SI
SE
RO
PT
PL
NL
LV
LU
LT
IT
IE
HU
GR
FR
FI
ES
EE
DK
DE
CZ
BG
BE
AT
Countries
Minimum and maximum average annual flood risk across
European countries measured in percent of GDP
 Regions and countries in Eastern Europe are particularly flood risk
hotspots
Managing flood risk on the European Scale: EUSF
Criteria for
funding met
Input: Loss distribution
Country A
Extreme value
distribution
estimated
Sampling
losses
…
Threshold
criteria
EUSF
payments
Sampling
Country Z
losses
Payment distribution for the EUSF
1
0.9
0.8
0.7
F
0.6
upper scenario
baseline
lower scenario
0.5
0.4
0.3
0.2
0.1
0
0
500
1000
1500
Million Euro
2000
2500
On average, every 7 years one can
expect that the EUSF can not
meet its obligations.
Managing current flood risk on the Country Scale
Government is a key actor:
Liabilities
Direct: obligation in any event
Explicit
Government liability
recognized by law or
contract
Foreign and domestic sovereign
borrowing, Expenditures by
budget law and budget
expenditures
Contingent: obligation if a particular
event occurs
State guarantees for nonsovereign
borrowing and public and private
sector entities, reconstruction of
public infrastructure
Implicit
A "moral" obligation
of the government
Future recurrent costs of public
investment projects, pension and
health care expenditure
Default of subnational government and
public or private entities, disaster
relief
Source: Modified after Schick and Polackova Brixi, 2004
Managing current flood risk on the Country Scale
Private sector and net loss
Government
Insurance
100%
90%
80%
70%
44%
58%
48%
62%
79%
60%
50%
40%
48%
30%
20%
17%
41%
11%
10%
0%
32%
1%
8%
Umbria EQ 1997
Poland Floods 1997
20%
21%
Austria Flooding
2002
Spain drought 2005
10%
Portugal drought
2005
Cross-country sample of financing modalities of disaster losses
by insurance, government assistance, and private sector and
net loss (as a percentage of direct losses)
Modeling Impacts and Adaptation: Country scale
Hazard
Floods, Droughts
Exposure
Capital stock, population
Direct Risk
Probabilistic asset losses
Economic resilience
•Financial resilience
•Economic redundancy
Economic vulnerability
Ability to recover and refinance
from disaster events
Indirect risk
Probabilistic fiscal and
Macroeconomic impacts
Physical Vulnerability
Susceptibility to
physical damage
Risk Management/
Adaptation
Development of risk
management strategies
Dynamic Modeling of Impacts and Adaptation: Country scale
Climate Change
Global Change
Ex-post
Hazard
Sudden onset
Dynamic
direct
risk assessment
Exposure
Physical
Vulnerability
Direct Risk
Economic
Resilience
Economic Vulnerability
Adaptation
and Risk Management
Economic Risk
Ex-ante
Dynamic
Indirect risk
management
Future Floods
Climate models remain limited
in their reproduction of local
weather extremes due,
inter alia, to inadequate (coarse)
resolution
Projected change in flood damages in
2071-2100 (% change with respect to
1961-1990 baseline)
Projections of changes in future
extreme weather events remain
highly uncertain and hinder us
from robustly projecting future
flood risk
Land use changes very difficult
to project
Managing current flood risk on the Country Scale
Need for managing risk on the country level: Hidden government
disaster liabilities
Government flood risk liability
Projected fiscal deficit 2009
6%
5%
Per cent of GDP
4%
3%
2%
1%
Lithuania
Slovakia
Bulgaria
Romania
Poland
Latvia
Hungary
Czech Republic
Austria
0%
Government fiscal deficits and hidden liabilities due to flood
risk in flood prone European countries
Austria case: Reserve Fund as a Risk instrument
• The Austrian “Katastrophenfond” (National Disaster Fund):
Reserve Fund Accumulation
500
400
Million Euros
300
200
100
0
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004
-100
-200
-300
Year
*
The Fund became negative in 2002 and 2003 (flooding).
To adjust the fund, investments more than 137 million in 2002 and
more than 207 million Euros in 2003 were needed.
* Source:Hyll et al., 2004
Austria case: Changes in fiscal space
Fiscal space concept: Fiscal flexibility is decreased due to disaster events
Austria fiscal space (bn Euro 2009)
25.0
20.0
Billion Euro 2009
15.0
Absolute w/o disaster risk
reduction (expectation)
reduction (standard deviation
reduction (95% quantile)
reduction( 99% quantile)
10.0
5.0
0.0
-5.0
-10.0
Stochastic trajectories of discretionary spending including disaster risk
Fiscal space and disaster related reduction in Austria
Time period of 10 years, 2009-2018
Drought
Approach used:
A
RCM
(SCENARIO a-A2)
Hazard
Exposure
Extreme events
frequency
LAND USE
(Corine)
CROP MODELLING
(Cropsyst)
B
Potential yield
Phenology
Actual yield
Yield loss
C
Yield loss return
period
Vulnerability
D
Current drought and heatwave risk
Losses in 2009 € millions
Similar analyses for
winter wheat, soybean
and sunflower
Annualised monetary risk due to combined heatwave and drought stress for
spring wheat calculated for the present period (1975-2005)
Future drought and heatwave risk
Changes in annualised drought and heatwave risks to spring wheat over a future period
in 2030-2060 based on SRES A2 (=2 degree) compared to today (in € millions)
Change in 2009 € millions
with adaptation: longer cycle variety
without adaptation
with adaptation: advanced sowing
Drought and heatwave:
•
With regard to drought and heat stress to agriculture, we find Southern
Europe to be particularly vulnerable
•
In a future climate with a north-south precipitation change gradient, and
assuming adaptation, many agricultural regions in Europe would actually
benefit from a warming climate
•
However, some regions in Italy and Spain would not be able to benefit and
adapt, and face continued stress and substantial associated risks
•
Drought and heatwave stress operate as slower onset phenomena and are
more strongly characterised by mean climate conditions: greater confidence
in model projections
Conclusions
•
Our study suggests that regions in Eastern Europe represent disaster
hotspots for flood risk, and areas in Southern Europe for drought and heat
stress to agriculture: case for increased cohesion funding?
•
Flood hazards are likely to worsen over much of Europe, yet due to a lack of
localised projections from climate models, we considered risk projections
not robust
•
In contrast, we feel more confident in projecting drought and heatwave risk
as well as adaptation as a function of changes in broader-scale average
climates
Conclusions
•
Although drought and heatwaves are likely to worsen across much of
Europe, effective adaptation interventions exist
•
Yet regional heterogeneity in risk and response will continue, leading to
climate change “winners” and “losers”.
•
Irrespective of future changes, weather-related disasters today already pose
substantial burdens for households, businesses, and governments
•
Risk-based adaptation planning seems important: prevention better than the
cure
End of Presentation
Partners
Introduction: Country perspective: Austria
• The Austrian “Katastrophenfond” (National Disaster Fund):
Reserve Fund Accumulation
500
400
Million Euros
300
200
100
0
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004
-100
-200
-300
Year
*
The Fund became negative in 2002 and 2003 (flooding).
To adjust the fund, investments more than 137 million in 2002 and
more than 207 million Euros in 2003 were needed.
* Source:Hyll et al., 2004
Introduction: European perspective: Solidarity Fund
• Fund may be heavily exposed to one (large scale) event only,
as experienced in 2002 with ¾ of the fund depleted due to flood events
Year
2002
2003
2004
2005
Aid granted (mill. Euro)
728
107
20
204
Successful Applications
4/4
6/10
1/11
10/12
• Nearly all (13/16) of the rejected applications were regional disaster
events.
• Given the fact that a number of new member states are very
hazard-prone and disaster losses are likely to rise,
the Solidarity Fund is likely to be severely under-funded in the future
Question: - How one could model disaster impacts (direct and indirect)
on the Country or European scale
- How to incorporate adaptation strategies
Modeling Impacts and Adaptation: Introduction
Risk bearers (aggregate country level):
- Public sector: Government: Infrastructure and Relief
- Private sector: business and households.
-property owners,
-insurers,
-reinsurers
-and the capital market.
Each stakeholder may implement a wide range of risk management
and adaptation strategies, including
- risk reduction: Structural or physical mitigation
- risk preparedness: Loss absorption, e.g. via savings
- and risk transfer: Risk spreading or financing
Furthermore, one can distinguish between ex-post or after-the-fact
approaches, and proactive (ex-ante) approaches.
Modeling Impacts and Adaptation: Direct Risk
Standard Approach: Four basic components
•
•
•
•
Hazard : Characterization of risk
Inventory: The elements at risk
Vulnerability: Susceptibility to damage
Loss: Direct or indirect
Structure:
Model output:
-
Hazard maps
Exceedance Probability (EP) curves
Probable maximum loss (PML)
Distribution of losses.
Modeling Impacts and Adaptation: Direct Risk
Exceedance probability curve
Loss frequency distribution
Direct Risk
Modeling Impacts and Adaptation: Indirect Risks
It is important to incorporate indirect effects within a risk
management framework
Possible indirect effects
on the macro-level
Economic vulnerability,
e.g. the ability to finance the losses,
is time dependent
Risk management/ adaptation strategies on the country level, have to
incorporate indirect risks as well in their decision strategy.
Modeling Impacts and Adaptation: Resilience
Public sector ex-post and ex-ante financing sources
Type
Source
Model
Ex-post sources
Decreasing government expenditures
Diversion from budget
Raising government revenues
Taxation
-
Deficit financing
Domestic
Central Bank credit
-
Foreign reserves
-
Deficit financing
External
yes
Domestic bonds and credit
yes
International borrowing
yes
Outside support,
yes
Ex-ante sources
Insurance
yes
Reserve fund
yes
Contingent credit
yes
Modeling Impacts and Adaptation: Ex-ante Options
Catastrophe Reserve Fund
Catastrophe Contingent Credit
Reinsurance
Mitigation
Modeling Impacts and Adaptation: Summary
Direct Risk: Probability of Asset losses (in monetary terms)
Economic Vulnerability: Ability to finance the losses
Economic Vulnerability is a function of
- Economic resilience: Loss Financing Instruments
- Direct Risk
Possible risk measures: Financing gap approach
e.g., shortfall between losses and financing possibilities
Indirect Risk: Probabilistic impacts and economic vulnerability
lead ultimately to macroeconomic effects.
Adaptation Strategies: Increase the economic resilience or
will decrease the direct risk.
Possible Hotspots in the European Union: Flood
Use direct risk (annual average losses) and debt indicators as
first proxies for economic vulnerability of the given country:
2.5
Losses/Capital Stock AAD MIN
2
Losses/Capital Stock AAD MAX
1.5
1
0.5
0
AT BE BG CZ DE DK EE ES FI FR GR HR HU IE IT LT LU LV NL PL PT RO SE SI SK UK
This approach would lead to the following countries as possible
Hotspots: Bulgaria, Romania, Croatia, Hungary, Czech Republic,
Slovakia, and Poland
Possible Hotspots in the European Union: Flood
-30°
-20°
-10°
0°
10°
20°
30°
40°
50°
60°
70°
80°
80°
-40°
70°
70°
Arctic Ocean
Ba
ltic
Se
a
60°
60°
Arctic Circle
50°
50°
North Sea
Atlantic Ocean
an
sp i
Ca
Black Sea
40°
40°
a
Se
Mediterranean Sea
-10°
0°
Legend
country
hotspot
0
1
:
10°
20°
Europe
30°
40°
50°
0
120
240
480
60°
720
960
Miles
Robinson Projection
Central Meridian: 30.00
Solidarity Fund: 0.4 billion Euros needed each year with a standard
deviation of around 0.3 billion (large scale events not incorporated)
Country Perspective: Austria Flooding
Direct Risk:
Loss Exceedance Curve
Probability of Exceedance
0.14
0.12
0.10
0.08
0.06
0.04
0.02
0.00
0
0.013
0.074
0.15
0.343
Losses in percent of Capital Stock
Economic Resilience:
• 2.5 percent of the direct losses of the public sector are financed from
outside assistance, namely the European Solidarity fund for losses
higher than 3 billion Euros or 0.6 percent of GDP.
• The disaster fund is set to 30 million Euros each year.
• The credit buffer, e.g. the maximum amount of credit from abroad the
government can or may use is set to 5 billion Euro.
Country Perspective: Austria Flooding
Ability to start new projects
Risk
Time
Fiscal consequences due to flooding in the next 10 years
Modeling Impacts and Adaptation: Direct Risk
Standard Approach:
Einführung in die Problematik: Effekte auf Landesebene
Oben: Mögliche
Entwicklungslinien
nach einer Katastrophe
Links: Durchschnittliche
Wachstumsraten nach
einer Katastrophe
Source: Hochrainer, 2006.
I:
Operationalisation of economic vulnerability
Description
Operationalisation in model
Financial
vulnerability
Ability to share risks
Financial vulnerability algorithm for
determining internal and external
savings for reconstruction, relief and
recovery given direct disaster losses
Economic
redundancy
Ability to pool risks: geographic
and economic diversification
CES function specification: input
factors are not perfectly substitutable
(not fully implemented )
Element
I:
Public sector ex post and ex ante financing sources for relief and reconstruction
Source
Considered in model
Decreasing government
expenditures
Diversion from budget
yes
Raising
revenues
Taxation
-
Central Bank credit
-
Foreign reserves
-
Domestic bonds and credit
yes
International borrowing
yes
Outside support, e.g. from EU
solidarity funds
yes
Insurance
yes
Reserve fund
yes
Contingent credit
yes
Type
Ex-post sources
Deficit financing
Domestic
Deficit financing
External
government
Ex-ante sources
CatSim: Software und Algorithmusstruktur
Financial vulnerability
Loss function: financing needs
10 year event
Probability
0.1
0.08
0.06
0.04
100 year event
0.02
200 year event
0
0
2,000
4,000
6,000
Losses
8,000
10,000
Financing sources: financing supply
12
International bonds
Marginal cost
10
Borrowing
from IFIs
8
financing
gap
6
Domestic
bonds
Diversion and credit
4
2
Grants
0
Amount available
12,000
Hazard
Floods, Droughts
Exposure
Capital stock, population
Direct Risk
Probabilistic asset losses
Economic resilience
•Financial resilience
•Economic redundancy
Economic vulnerability
Ability to recover and refinance
from disaster events
Indirect risk
Probabilistic fiscal and
Macroeconomic impacts
Physical Vulnerability
Susceptibility to
physical damage
Risk Management/
Adaptation
Development of risk
management strategies