CAPRA - Probabilistic Risk Assessment Initiative (Manzul Hazarika)

Disaster Risk AssessmeDisaster Risk AssessmeApplications

Manzul Kumar Haz

With contrib

Gabriel BernaFrancis Ghesquiere, T

ent and Management and ent and Management and s of CAPRA

arika, AIT, Thailand

butions from

l, UPC, SpainThe World Bank, USA

Defining ‘Disaster Risk’

Potential Hazard

Hazard cause dand theiHazard• Hazar• It is re• It is va

V l bl

• The ev

Vulnerable

Risk ~ Hazard → StruRisk Hazard → StruRisk ~ Hazard x Vulnerability → QuaRisk ~ Hazard x Vulnerability/Capacity → Qua

(H) is a phenomenon or situation, which has the potential to isruption or damage to people, their property, their services r environment.

rd is a probabilityestricted for a given period of timealid for a specified area

A

vent has a certain intensity – or magnitude

e Area Elements at Risk

uctural Measuresuctural Measuresantitative Risk Assessmentalitative Risk Assessment, CBDRM

Risk is Also a Spatial Pro

Hazard: How big is the flood (10 yr 50 yr 100 yrHazard: How big is the flood (10 yr., 50 yr. 100 yrVulnerability: What is the depth of water in elemeElements at risk: Which elements are getting aff

oblem

r )?r.)?ents at risk?fected?

How to Assess Risk Spatially

Risk = Hazard x Physical Vu(Probability of occurrence) (Degree of losses

ge (%

)

VulnerHazard

Flood Depth

Dam

ag5 1025 1005 1025 100

Type of hazardIntensityDurationS ti l E t t ExposureSpatial Extent

Overlay of hazard

Exposure

y?

ulnerability x Amount/Numbers to elements at risk) (Quantification of exposed elements)

rability Elements at risk

T f l t t or Intensity

Type of elements at riskNumbersEconomic valueL tiLocation

d & element at risk

Importance of Disaster Risk

Disaster risk assessment important for deveRisk assessment provides a good platform fEvidence provided by risk assessments helpRisk assessment along with cost-benefit anaRi k A t b i i i dRisk Assessments are being commissioned

ChallengesSignificant increase in scientifically-based abut without much convergence;Limited collaborations among the various plLimited collaborations among the various plRisk assessment need variety of data, but d

5

k Assessment

elopment and risk reduction decision making;for advocacy with the policy or decision maker;ps to get political support and mobilize resources;alysis can help in prioritizing DRR investments; i iti (D lhi A Dh k t )in many cities (Delhi, Amman, Dhaka etc.).

nd community-based (VCA) risk assessments,

ayers involved in risk assessments;ayers involved in risk assessments;data sharing/inter-operability remains a challenge.

Disaster Risk is Increasing

Main driver of disaster risk is the rapidly incrPoorer countries have disproportionately hig

Rank

FloodAbsolute Relative Absolute

Absolute and Relative Exposure of Top Asi

R (Million) (%) (Million)1 Bangladesh1 (19.2) Cambodia1 (12.2) Japan1 (30.9

2 India2 (15.8) Bangladesh2 (12.1) Philippines2(20.7

3 China3 (3.9) Vietnam3 (3.9) China3(11.1

4 Vietnam4 (3.4) Bhutan4 (1.7) India4(10.7

5 Cambodia5 (1.7) India5 (1.4) Bangladesh6(7.5

6

reasing exposure;gher mortality/GDP risk.

Cyclone EarthquakeRelative Absolute Relative

ian Countries:

(%) (Million) (%)9) North Marina Isl.2 (58.2) Japan1(13.4) Vanuatu1 (60.4)

7) Niue9 (25.4) Philippines2(12.1) Solomon Isl.2 (36.3)

) Japan10 (24.2) Indonesia3 (11.0) Tonga6 (21.1)

7) Philippines11 (23.6) China4 (8.1) Papua New G..9 (17.5)

5) Fiji12 (23.1) India8 (3.3) Philippines12 (13.8)

Source: Asia Pacific Disaster Report, 2010

HFA and Disaster Risk Asses

HFA Priority Actions ActivitiesAction 1: Ensure that – Legislation to DRR is a priority support DRRAction 2: Identify, assess, and monitor

– Risk assessmen– Early warning,

disaster risksy g

Action 3: Knowledge & education to build a

– Research in mulhazard riskeducation to build a

culture of safetyhazard risk

– Public awarenesAction 4: Reduce

d l i i k f t– Land-use plannin

B ildi l tiunderlying risk factors – Building regulatioAction 5: Preparedness for effective Response

– Review/exercisepreparedness anp p pcontingency plan

7

ssment/Disaster Management

Challenges Reported– Political inertia in approving legislations

nts – Lack of technical/financial capacities– Difficulty in early warning at very local levely y g y

ti- – Multi-hazard risk ass. remains a weak areaGap in risk communication for public

ss– Gap in risk communication for public

awarenessng – Weak enforcement of land-use planning

Diffi lti i i l ti b ildi don – Difficulties in implementing building codese nd

– Eighty percent countries have contingency plans to deal with major disasters

nsp j

Framework for Disaster Risk AManagement

Hazard ITriggering Factors Environmental Factors

GeologySoilLand-use

RainfallVolcanic EruptionEarthquake

LandslidFloodDrought

TopographyHydrology

qCycloneEarthqua

Spatial-Temporal Probability

Risk = Hazard

Quantitative Risk Assessment (Absolute Risk)

Disaster Ma

Disaster PrDisaster Mitigation (Disaster Risk Management)

8

Assessment and Disaster Disinventar

nventory Elements at Risk

de BuildingsInfrastructuresPopulation

Socio-Economic Factors

LivelihoodsLiteracyGender

akeCritical FacilitiesLifelines

PovertyCulture

Community Approach

S l C l d

x Vulnerability

Damage-intensity Relationships Seasonal CalendarTransect Walk

Qualitative Risk Assessment (Relative Risk)

anagement

reparedness Disaster Response Disaster Recovery

Hazard Assessment by Mo

Historical river discharge record was analyzed to find th f it d

4000

5000

6000

umec

s)

the frequency-magnituderelationship for flood

0

1000

2000

3000

Disc

harg

e (cu

01.00

20 Year

Flood hazard maps obta9

odeling

Discharge vs Return Period

Modeling10.00 100.00

Return period (years)

50 Year 100 Year

ined from a flood model (Kalu-Ganga Basin, Sri Lanka)

Hazard Assessment by ParApproach

Flood Haz20 years 20 years 50 years

100 years 10

rticipatory/Community

zard Assessmentflood - Depth up to 90 cmflood Depth up to 90 cmflood - Depth up to 250 cmflood - Depth up to 400 cm

Vulnerability Assessment

Types of Vulnerability

1) Physical – e.g., types and locations of buildings

2) Economic – e.g., non-diversified livelihoods, pove

3) E i t l/ t l 3) Environmental/natural – e.g., destruction of natural barriers e

4) Individual 4) Individual – e.g., age, gender, skill etc.)

5) Social l d hi t )– e.g., poor leadership etc.)

11

Qualitative Vulnerability

Quantitative Vulnerability

Detail Survey

(S i tifi )

erty etc.)

(Scientific)

etc.)Questionnaire Survey

(Traditional VCA)

Quantitative Vulnerability

12

Assessment

Qualitative Vulnerability Ass

Sensitivity Demogr

Major CoContributing Factors

Sensitivity g

LVULNERABILITY

Rural

Adaptive CapacityAdaptive CapacityExposure

Flood Depth

Duration of Flood Events

Exposure: Intensity and duration of flood eventsAd ti C it C it ’d bilit t ith t d Adaptive Capacity: Community’d ability to withstand orSensitivity: Degree to which the community is affected

13

sessment for Flood

raphic Standing Gender

omponents Sub-Components

p g

Land

Gender

Age Groups

Health ConditionStanding

Water

Health Conditiong

Income

LivelihoodMajor Components

Sub-Components

Assets

f fl d t

j p

Contributing Factorsr recover from flood events by flood events.

Hanh et al., 2009

Field Data Collection for QuaAssessment

Contributing Factor Major Components Sub-C

Sensitivity (S) Demographic standing Gender, Age, HRural standing Accessibility, pWater Well, bore-welLand Ownership, siz

Adaptive Capacity (A) Income Amount sourcAdaptive Capacity (A) Income Amount, sourcAssets House types/sLivelihoods Agriculture, bu

Exposure (E) Past flood events Flood frequenc

14

alitative Vulnerability

Components

Healthproximity to marketll, tap waterze, land-usecesces

size, vehicleusiness, jobscy, flood depth

Qualitative Flood Vulnerabil

S (HH)

FVI=0.1FVI=0.5

Flood Vulnera

FVI=0.FVI=0 5FVI 0.5

FVI=0.8FVI=0.5FVI=0.8

E (HH) A (HH)

15

ity Map

4815047

ability Index Sensitivity (S) Adaptive Capacity (A) Exposure (E)

1481 0.4575 0.5749 0.2402 5047 0 4711 0 3235 0 4355 5047

84905047 0.4711 0.3235 0.4355 8490 0.6222 0.3158 0.7201

Quantitative (Absolute) Ri

Risk = Hazard x Physical Vulnerability x Amoun

Hazard (H) = Probability of occurrence Vulnerability (V) = Degree of loss to a given Amount (A) = Value of the element (s) a

US $ 50,000

US $ 200,000

V = 0.6 $ ,

V = 1

Risk = Hazard x Vulnerability x Amount= 0.1 * ((0.6*200,000) + (1 * 50,000) + (0.1*10 0.1 ((0.6 200,000) (1 50,000) (0.1 10= 0.1 * 180,000 = 16,000 $

16

sk Assessment

nt

of a hazard (0 to 1) element (s) at risk, 0 (no loss) to 1 (total loss).at risk

US $ 100,000

10 years FloodV = 0.1

00,000) ) 00,000) )

Quantitative Risk Map for

Quantitative VulnHazard

x

Flood Risk Map

17

Flood in Sri Lanka

nerability

Amount of loss for Amount of loss for various types of

buildingsx

Qualitative Risk Map for F

Hazard

x

18

Flood in Sri Lanka

Qualitative Vulnerability

COMPREHEN

CAPRA was

Consortiu

19

SIVE APPROACH FOR PROBABILISTIC RISK ASSESSMENT

s developed by,p y

um ERN-AL

CAPRA InitiativeRisk analysis methodology

PHazardD

VulnerabilityVulnerability

Exposure

20

Loss EstimationPhysical oss st at oEconomic HumanDamage

Application Application 11Application Application 11

A li ti A li ti 11

Applications

Application Application 11Application Application 11

CAPRA InitiativeObjective and scope

Development of tools forcommunication in order to:

Sensitize decision makers abonatural disasters;natural disasters;

Formulate risk management stand sub-national levels;;

Develop a common methodorisk

21

r risk assessment and

out the damaging potential of

trategies at regional, national

logy for quantifying disaster

CAPRA InitiativeObjective and scope

The CAPRA initiative aims toregional strategy versatileregional strategy, versatiledevelopment of risk assessmerelated to management of disast

22

become the focus of aand effective for theand effective, for theent and decision makingter risks.

CAPRA InitiativeFeatures

Probabilistic risk modeling

Visualization moduleVisualization module 

Open architecture platform

23

Hazard modules 

Exposure and vulnerability modules 

Risk assessment module 

Variable resolution grids

Visualization tools 

Planning tools 

GIS interaction

Open sourceOpen source

User oriented applications 

User updatable

CAPRA InitiativeFeatures

Oriented software

W b itWebsite

Communication strategy

24

Available hazard mapsp

Some exposure databases

Vulnerability curves library

Risk and loss analysisRisk and loss analysis

Information datasets

S ft d l dSoftware download

Wiki

Open access trough internet

Regional seminars and workshops

Books and papersBooks and papers

CAPRA InitiativeWhy a probabilistic risk analy

The local intensities produced boccurrence frequencyoccurrence frequency

There is uncertainty in the estimatio

Risk must be expressed in terms of

Abili dAbility to compare and aggregateor multi-risk

Likely future scenarios (climate cha

25

ysis?

by dangerous events (hazard) have an

on of hazard and vulnerability

f occurrence rates or return periods

l f diff l i h dlosses of different events– multi-hazard

nge)

25

CAPRA Players

Comprehensive Approach for Pr26 robabilistic Risk Assessment (CAPRA)

Selected CAPRA ResulManagua (Nicaragua) earthquake r

Economic loss for 6.5 Mw earthquake occurrin

27

lts isk

ng near Managua

AAL/Building

Results for NicaragBluefields hurricane risk

Relative economic loss. Probabilistic analysis f

28

gua

for Wind only

AAL/Eco. Value (Building)

Results for NicaragCorinto hurricane risk

Relative economic loss. Probabilistic analysis f

29

gua

for Wind and storm surge combined

AAL/Eco. Value (Building)

Results for NicaragCorinto hurricane risk

Human expected loss. Probabilistic analysis fo

30

gua

or Wind and storm surge combined

Results for Costa San Jose earthquake risk

Relative economic loss

31

Ricak

AAL/Eco. Value (Homogeneous zone)

Results for Costa San Jose earthquake risk

Human expected loss

32

Ricak

Results for El SalvSan Salvador multi-hazar

Annual average loss (relative). Probabilistic an

33

adorrd risk

nalysis (earthquake and hurricane)

Results for El SalvSan Salvador multi-hazar

Annual average loss (absolute). Probabilistic a

34

adorrd risk

analysis (earthquake and hurricane)

Results for BelizeBelmopan earthquake ris

Annual average loss (absolute)

[US$]0 - 31

32 - 10532 105

106 - 253

254 - 554

555 - 1,421

35

sk

Results for BelizeBelmopan hurricane (win

Annual average loss (absolute)

[US$]0 - 222

223 - 423

424 - 794

795 - 1,506

1,507 - 5,644

36

nd) risk

Results for BelizeBelize City earthquake ri

Annual average loss (absolute)

[US$]7 - 603

604 - 1,453

1,454 - 2,941

2,942 - 7,500

7,501 - 24,722

37

sk

Results for BelizeBelize City hurricane (win

Annual average loss (absolute)

38

nd and storm surge) risk

[US$]472 - 39,077

39,078 - 95,905

95,906 - 185,579

185,580 - 331,062

331,063 - 765,792

Risk Analysis AppPossible Applications of CAP

1) Visualization of hazard and )risk

3) Infrastructure design

5) Contingency and Emergency Planningg y g

7) Real time post-event damage estimation

39

licationsPRA

2) Land use and territorial )planning

4) C/B analysis and support of risk reduction measures

6) Damage scenario analysis

8) Design of financial protection instruments

CAPRA Applications in Bo

40

ogotá

• Colombia’s capital and main economic center

• 750 000 residential buildings750,000 residential buildings

• 200 hospitals and health centers

• 3 500 schools and Universities3,500 schools and Universities

• Population: Over 7,000,000 (High concentration of pop.)

• Exposed value US$40 Billion only building constructions + infrastructure

1) Visualization - Hazard

Seismic Lands

41

d Mapping

slides Floods

2) Landuse and Territor

42

ial Planning

42

43

Before…Resettlement from Nueva Esperanza Barrio

44

… afterResettlement from Nueva Esperanza Barrio

3) Infrastructure Desig

45

gn (Specifications)

Colegio Distrital Florida Blanca

46

4) Cost Benefit Analysis

47 Source: GAR, 2011

etrofitting of Critical Infrastructur

48

re in Bogota

6) Damage Scenario Anal

DAY

49

lysis

NIGHT

6) Contingency and Emer

• Location of emergency units• Functional vulnerability (emergency rou• Health services requirements• Housing requirements• Food requirementsFood requirements• Utilities requirements (water, energy, e• Debris and construction materials

50

rgency Planning

utes, etc.)

tc.)

7) Immediate Damage Es

Bogotá Accelerograph Network (RA

51

stimation

AB)

Immediate Damage Estim

DamSurface Response Spatial Distributionp

52

mation

Physical damage

Human losses

mage Distribution Calculation

8) Design of Financial Pro

Technical premiumTechnical premium(annual expected loss)

53

otection Instruments

Insurance premium calculationInsurance premium calculation

Conclusions

One risk model (CAPRA) can be used

Cost of modeling is going down- geospatial technologies- open source models- open source models

“Models remain models”

54

d for a large number of applications

Upcoming Training on CAPRA

55

A

Date: 28 Nov – 02 Dec, 2011Location: AIT, Bangkok

SParticipants: From South Asian countriesFees: Free! Travel/Accommodation: Borne by participants/ y p p s

www.ecapra.org

Th k f Thank you for yo ki d tt tiour kind attention

Generation of the Risk Cuurve

Need at least 6-7 scenarios to develop a risk curve

Area under this curve shows the average annual loss