Toshio OKAZUMI
Flood Risk Assessment and Disaster Data - Users’ perspective -
Chief Researcher International Centre for Water Hazards and Risk Management
(ICHARM) /Public Works Research Institute (PWRI)
October 1, 2013 “Expert Group Meeting on Improving Disaster Data
to Build Resilience in Asia and Pacific” Sendai, Japan
6 March, 2006 at Tsukuba
ICHARM
International Centre for
Water Hazard and Risk
Management under the auspices of UNESCO hosted
by PWRI, Tsukuba
Objective: To serve as the Global Center of Excellence to provide and assist implementation of best practicable strategies to localities, nations, regions and the world to manage the risk of water related hazards including floods, droughts, land slides, debris flows and water contamination
Advanced Early Warning System IFAS
ICHARM’s Policy : Localism Delivering best available knowledge to local practices
Local Practices
Integrated
Flood Analysis
System (IFAS)
Flood risk assessment
under climate change
Supported by MEXT
Master Course
Hazard Mapping Course
River & Dam Course
IFI
WWAP, AWDO
IFNet/GFAS
Sentinel Asia
Flood Preparedness
Indicators/Standard
WWF, APWF
The 1st Phase
Focus: Flood-related
Risk Management
Education Research
Ph.D. Course
Supported
by JICA
UNESCO Centers
IRDR
UNISDR GP-GAR
Working as a Knowledge Hub on WD through RETA
Supported by ADB UNESCO Pakistan pjt
Users’ Perspectives on Disaster Data
1. Definition of Terminology
2. ICHARM Challenges
a. Global Risk Indices
b. Flood Risk Assessment in River Basin
3. Conclusions
1. Definition of Terminology
Risk = Hazard × Vulnerability
Risk
Avoidance
(moving to safer location, etc.) is one good measure of disaster management
Hazard
Community/ Society
Hazard
Community/ Society
However, we cannot avoid hazards to come in most cases
The First Capacity: Prevention : Hazard cannot enter because of Preventive measures
Coping Capacity
Prevention
Hazard
Community/ Society
Hazard
Community/ Society
Hazard
Community/ Society
The Second Capacity: Coping Capacity: Hazard can enter Vulnerable situation but coping capacity will reduce the damage from hazard
Definition of terminology
8
Words Definitions Remarks
Vulnerability Amount of Potential Damage caused to a system by a particular climate-related event or hazard.(Jones and Boer 2003)
Unit is monetary value
Hazard A dangerous phenomenon, substance, human activity or condition that may cause loss of life, injury or other health impacts, property damage, loss of livelihoods and services, social and economic disruption, or environmental damage.
Non-Dimension (Probability)
Risk The combination of the probability of an event and its negative consequences.
Unit is monetary value
9
Risk = f ( HAZARD , Vulnerability ) = f ( HAZARD , Exposure, Basic Vulnerability, Capacity) Indicator of extreme statistics of hazard type
Affected by particular event (e.g. Population and property)
• Exposure in Flood Inundation Area To identify sensitivity of Exposure (E) to change flood prone area (H)
A potential social factor which causes exacerbation of the damage
• Disaster damage Not only hazards but also vulnerability, the human side of disaster causes, should be expected per area (maximum loss of persons and asset)
H E
2. ICHARM Challenges
a. Development of Global Risk Indices
I. BASIC PHILOSOPHIES
07/10/2013 11
There are a lot of questions in the existing global risk
assessment methods.
- Most of all, risk is estimated by the conceptual structure, and as a result, indices are far from real phenomena.
- Meanwhile, the upcoming global disaster risk reduction framework (HFA2) will require “evidence-based risk information”.
To explain real phenomena, the new methodology for risk
indices should be developed in the following manners:
- Upscale localized or river-basin leveled methods in assessing risks - Simplify those methods with consideration of data availability - Develop global risk indices with good credibility
RISK
ASSESSMENT
07/10/2013 12
II. OUTLINE For a comprehensive, integrated, multi-disciplinary approach
Quantifying our present risks
Preparedness
Assessment
Vulnerability
Assessment
Quantifying natural aspects of present risks III
IV
V
HAZARD
ASSESSMENT
I • Output: how hazardous are natural conditions?
• Example: flood inundation depth in an area
EXPOSURE
ASSESSMENT
II • Output: how many people or assets are affected by hazards?
• Example: number of people in hazardous areas
• Output: what factors increase or decrease potential damage? - identification of factors which influence risk
• Example: proxies related to marginalized group, unplanned urbanization, political corruption, and coping capacity • Output: how people or our assets
would be damaged?
• Example: fatalities, asset losses, and productivity losses
• Output: evaluation of effectiveness of disaster risk reduction actions (including structural & non-structural measures)
• Example: how effectively DRR education and early warning systems would reduce the present risks
Actions & well-
designed monitoring
activities
Global BTOP model of ICHARM
1. BTOP model – schematic diagrams 2. DEM upscaling (60N-60S)
Method descriptions
- Block TOP model with Muscingum-Cunge river discharge routing
- Cell size: 20 km losses DEM up-scaling
Advantages and possibilities
- Distributed hydrological model
- Runs on individual river basin
- Global application with up to 1 km
- Poorly gauged basins with satellite precipitation data
APPLICATION: Daily River Discharge
Green : 1km to 20km Blue : 90m to 20km
PROTOTYPICAL OUTCOMES OF ICHARM (1)
PROTOTYPICAL OUTCOMES OF ICHARM (2)
FID simulation
Method descriptions - Topographical algorithm for computing the
potential flood inundation depth
Advantages and possibilities - High resolution (500m ~1km) in global scale
and run time is fast - Globally acceptable performances - GIS-based approach, encouraging spatial
risk assessment
Condition
50-year return period discharges based on the
BTOP model From MRI-GCM3.2S BC (1980-2004)
(m3/s)
SPAHigh : 500000
Low : 0
Present_peak_discharge
SRTM DEM(HydroSHED15S)
Global FID
Shallow: 0.1 m
deep : over 10 m
FID =ƒ (DEM, Hmax, Flow direction, i ) Hmax = { Qmax * n0/(Bf*If)}^0.6
Input Data:
HydroSHED global-scale
Potential Flood inundation depth
PROTOTYPICAL OUTCOMES OF ICHARM(3)
1. Input conditions () : floodwater depth and population data2009 (1km)
Population distribution > 1 person/ km2
2. Results of exposure assessment 12S
47E 147E
56N
Affected people
Thin: 1 person/km2
Dense : over 1000
Coupling FID simulation and LandScanTM Method descriptions
- calculation of affected people by overlapping FID simulation results and high-resolution digital population map
Merits and advantages - reliable to estimate affected people in the
unit cell, because of high resolutions of FID and LandScanTM
- GIS-based approach, ensuring the further spatial risk assessment
1. Input conditions Flood hazard and exposure: FID simulation in case of 50-year return period discharge.
<ASIAN APPLICATION: 14 countries >
47E 147E
56N
12S
Affected people
Thin: 1 person/km2
Dense : over 1000
1. Study sites: 14 countries in Asia Flood hazard and exposure - simulated at the 50-year return period Proxies for vulnerability and risk calculation formula - estimated
2. Results of mapping flood risk Affected people: 1km cell-gridded Risk index (flood fatalities) was measured at country level, but it was possible to anticipate hotspots within the country.
<FLOOD RISK MAP> Kazakhstan RISK = 0.01
Bangladesh RISK = 0.16
Nepal RISK = 0.14
Pakistan RISK = 0.08
Cambodia RISK = 0.07
Viet Nam RISK = 0.06
Laos RISK = 0.04
Thailand RISK = 0.05
India RISK = 0.06
Indonesia RISK = 0.04
Korea RISK = 0.03
Japan RISK = 0.01
Philippines RISK = 0.06
Tajikistan RISK = 0.02
RISK = flood fatalities per 100 km2
07/10/2013 17
III. ADVANTAGES AND POTENTIALS IN ICHARM’S RISK ASSESSMENT METHODS
HAZARD ASSESSMENT - APPROACH: global BTOP, and FID simulation - ADVANTAGES: overcoming data unavailability of developing countries with simplified
hydrological models
EXPOSURE ASSESSMENT - APPROACH: GIS-based assessment through coupling of FID simulation and LandScanTM - ADVANTAGES: reliable to estimate affected people for each grid-cell
VULNERABILITY ASSESSMENT - APPROACH: selection of more direct proxies, and application of advanced statistical methods - ADVANTAGES: evidence on importance of various proxies at the disaster damage viewpoint
RISK INDEX CALCULATION - APPROACH: organization of all data into risk indices by using the risk calculation formula - ADVANTAGES: risk indices are not conceptual, and enable us to consider various hydro
extremes, and important social conditions
POTENTIAL AS A VEHICLE FOR POLICY-MAKERS AND DECISION-MAKERS - Providing risk information with “much higher credibility”
2. ICHARM Challenges
b. Flood Risk Assessment in River basin
Process of Flood Risk Assessment
1. Data collection and analysis on precipitation
2. Data Collection and analysis on information in river basin 3. Inundation simulation
Runoff calculation Hydraulic analysis Inundation analysis
4. Assessment of Hazard and Vulnerability ×
<Hazard> Analysis on Inundation area, depth, duration and reach time
<Exposure> Analysis on Vulnerability of Population, Assets, infrastructures, Core functions
5. Risk Assessment
<Assessment area> Economic Damage, Human Casualty, Impact and Damage to Urban function, Recovery, Indirect impact to private company
<Assessment method> Quantitative assessment Qualitative assessment
data
technology
(Source: River and Sabo technical guideline, MLIT-Japan)
20
Water depth / duration distribution
House damages
Damages calculations
Agricultural damages
Damage curves
House value distribution
Cultivation starting date
Results
Calculation
Example;
Uncertainty Estimation during Flood Risk Assessment in the
Pampanga River, the Philippines
21
22
Pantabangan dam
Arayat
San Isidro
Metro Manila
Pampanga river basin A= 10,434km2
Angat dam
1
10
100
0 1000 2000 3000 4000 5000
Ret
urn
Peri
od
Discharge(m3/s)
Gumbel
Gumbel
Arayat_revised
Arayat_original
so
Correction of observed data
Gap = 20%
San Antonio Swamp
Candaba Swamp
Max inundation depth in Sep 2011 flood
Results of Runoff model Flood simulation (flood in Sep 2011)
San Isidro
Mayapyap
(m)
(m)
Retarding effect of Swamp can be seen in hydrograph
Land Use in Pampanga river
Change of Land Use at San Isidro 24
1
10
100
0 1000 2000 3000 4000
Ret
urn
Peri
od
Discharge (m3/s)
Annual maxiumdaily dischargeGumbel
IFAS
BTOP sm
Uncertainty of runoff model selection
Gap = 40%
Result of Rainfall and Runoff Inundation(RRI) model simulation in Sep 2011 flood
Every hour from Sep 26 to Oct 4 (9 days)
Water Depth(m) 26
Growth stage
Days of submerge 1-2 3-4 5-6 7
Estimated yield loss (%) Vegetative stage: Minimum Tillering /Maximum Tillering
10-20 20-30 30-50 50-100
Reproductive Stage: Panicle Initiation/Booting Stage (Partially Inundated)
10-20 30-50 40-85 50-100
Reproductive Stage: Panicle Initiation/Booting Stage (Completely Inundated)
15-30 40-70 40-85 50-100
Maturity Stage: Flowering stage 15-30 40-70 50-90 60-100 Ripening Stage 5 10-20 15-30 15-30
Flood damage matrix: Rice Damage
0
20
40
60
80
100
1-2 3-4 5-6 7
Pe
rce
nta
ge o
f Y
ield
Lo
ss
(%)
Durations of Flood (Days)
Ripening
0
20
40
60
80
100
1-2 3-4 5-6 7
Pe
rce
nta
ge o
f Y
ield
Lo
ss
(%)
Durations of Flood (Days)
Maturity
0
20
40
60
80
100
1-2 3-4 5-6 7
Pe
rce
nta
ge o
f Y
ield
Lo
ss
(%)
Durations of Flood (Days)
Vegetative
0
20
40
60
80
100
1-2 3-4 5-6 7
Pe
rce
nta
ge o
f Y
ield
Lo
ss
(%)
Durations of Flood (Days)
Reproductive
Ranges of yield loss of rice due to flood
Drought damage matrix: Rice and Corn Damage
Growth stage
Estimated yield loss (%)
PALAY CORN Seedlings 40 25 Vegetative 50 50 Reproductive 60 80 Maturity 15 15
Agricultural Damage can be defined as:
= function ( flood duration, growth stage)
Risk Assessment: Agricultural Damage (Flood and Drought)
Maximum damage with location of Pampanga province and Calumpit Municipality
Minimum damage with location of Pampanga province and Calumpit Municipality
Max - Min Gap = 50%
Observed data Hazard Analysis
Flood Reproduction calculation
Risk Assessment
so sm sr
S
Field investigation Satellite info
Flood Mark Inundation map Satellite observation (MODIS, SAR, PRISM)
Damage data Damage function Satellite image
20
%
40
%
50
%
Uncertainty estimation during Flood Risk Assessment Uncertainty reduction method
Necessity data for Flood risk assessment
1. Hydrological data ; to identify hazard and its probability, validation of hydrological and hydraulic model ( professional agency)
2. National Statistics; to identify exposure with overlaid to No.1(population, asset, etc.)
( National Statistic Office)
3. Damage data; to identify risk (causality and economic losses), validation of risk model
( this should be enhanced)
Results of Risk Assessment will be utilized as decision-making info. Reduction of uncertainty as much as possible is important.
Thank you very much