Toshio KoikeDirector, International Centre for Water Hazard and Risk Management (ICHARM)

Professor Emeritus, the University of TokyoCouncil Member, Science Council of Japan (SCJ), Cabinet Office of Japan

Chair, River Bureau of Japan

Innovative Science and Technology for Reducing Water-related Disaster Risk

1

Building Resilience

Mar

ch 2

015 Sendai

Framework on Disaster Risk Reduction

Sept

embe

r 201

5 Sustainable Development Goals

Dec

embe

r 201

5 Paris Agreement(COP 21)

Concerted Actions are RequiredPreventing Future Risk

Reducing Current Risk

Adaptation& Recovery

Sustainable Development

Understanding Governance Investment EW/BBB

Key Global Agendas

Quantifying uncertainty

Climate models

Multi-model

ensemble (MME)

Down-scaling

Basin-scale

prediction of

quantity & quality

Water quantity

and quality prediction

flood

ordinary water

drought

ground water

Information

Storage

Treatment

Current facility, plan, management

Flood control system

Water allocation &

cost

Environment

Human life

Industry

HumanBehavior

EconomicBehavior

Drought Simulation

Flood SimulationIm

pact assessment

Filed survey

Early warning

Allocation policy

Land use

Adaptation options

Innovative technology- Flood control

- quality control

Decision m

aking

Monitoring evaluation

implem

entation

IntegratedObserved Data Sets

ProcessStudy

Scientific approach Engineering Approach Socio-economical approach

3

End to End Approach on Climate Change Adaptation

Quantifying uncertainty

Climate models

Multi-model

ensemble (MME)

Down-scaling

Basin-scale

prediction of

quantity & quality

Water quantity

and quality prediction

flood

ordinary water

drought

ground water

Information

Storage

Treatment

Current facility, plan, management

Flood control system

Water allocation &

cost

Environment

Human life

Industry

HumanBehavior

EconomicBehavior

Drought Simulation

Flood SimulationIm

pact assessment

Filed survey

Early warning

Allocation policy

Land use

Adaptation options

Innovative technology- Flood control

- quality control

Decision m

aking

Monitoring evaluation

implem

entation

IntegratedObserved Data Sets

ProcessStudy

4

Scientific approach Engineering Approach Socio-economical approach

End to End Approach on Climate Change Adaptation

MODEL SELECTION: Precipitation

(May-November)

Underestimation of extreme rainfallExtreme rainfall correction

too many low drizzle/ tiny rainfall No rainfall correction

Large Diversity Low Seasonal variation

Normal rainfall correction

Bias in GCM

GCM raw seasonality (1986-2000)

0

5

10

15

20

Jan

Feb

Mar Ap

r

May

Jun

Jul

Aug

Sep

Oct

Nov

Dec

rain

fall

mm

/day

Bandaunginmcm30csiro_mk3_5gfdl_cm20gfdl_cm21miroc32hire

Ranking Raw rainfall (1986-2000)

0

50

100

150

200

rain

fall

mm

/day

Bandaunginmcm30csiro_mk3_5gfdl_cm20gfdl_cm21miroc32hire

1500 2000 2500 3000 3500

2

4

6

8

10

12

14

16

18

20

22

rank

rain

fall

mm

/day

Long drizzle rainy day problem

Bandaunginmcm30csiromk35gfdlcm20gfdlcm20miroc32hire

7

Insitu-station Corrected GCMavg Future Corr_GCMavg Future - Past

Probability:10year

Probability:100year

Quantifying uncertainty

Climate models

Multi-model

ensemble (MME)

Down-scaling

Basin-scale

prediction of

quantity & quality

Water quantity

and quality prediction

flood

ordinary water

drought

ground water

Information

Storage

Treatment

Current facility, plan, management

Flood control system

Water allocation &

cost

Environment

Human life

Industry

HumanBehavior

EconomicBehavior

Drought Simulation

Flood SimulationIm

pact assessment

Filed survey

Early warning

Allocation policy

Land use

Adaptation options

Innovative technology- Flood control

- quality control

Decision m

aking

Monitoring evaluation

implem

entation

IntegratedObserved Data Sets

ProcessStudy

9

Scientific approach Engineering Approach Socio-economical approach

End to End Approach on Climate Change Adaptation

ICHARM Hydrological Simulation SystemIntegrated Flood Analysis System(IFAS) Rainfall-Runoff-Inundation (RRI)

Flood Hazard Analysis by RRI Simulation

Flood Hazard Assessment in Pampamga River Basin

Using IfSAR DEM

Inundated area (>0.5m depth)= 77,396 ha

Inundated area (>0.5m depth)= 103,376 ha

25-Year Flood 50-Year Flood 100-Year FloodInundated area (>0.5m depth)= 127,008 ha

Different Flood Scale

Maximum Inundation Depth

Interferometric Synthetic Aperture Radar (IfSAR) provided by National Mapping and Resource Information Authority (NAMRIA)

450m x 450m grid

11

ICHARM Hydrological Simulation SystemIntegrated Flood Analysis System(IFAS) Rainfall-Runoff-Inundation (RRI)

Ensemble Rainfall

Prediction

12

Water andEnergy Budget

DistributedHydrological

Model(WEB-DHM)

Energy-Water

Balance

River Flow

Soil MoistureDynamics

Lateral Flow

Flood

Obs. River Discharge

EnsembleMean

SinglePrediction

270

280

290

300

310

320

330

340

350

1-Jan 1-Feb 1-Mar 1-Apr 1-May 1-Jun 1-Jul 1-Aug 1-Sep 1-Oct 1-Nov 1-Dec

Wat

er L

evel

, m

Sameura Dam Water Level: Past

19811982198319841985198619871988198919901991199219931994199519961997199819992000Average

270

280

290

300

310

320

330

340

350

1-Jan 1-Feb 1-Mar 1-Apr 1-May 1-Jun 1-Jul 1-Aug 1-Sep 1-Oct 1-Nov 1-Dec

Wat

er L

evel

, m

Sameura Dam Water Level: Future

20462047204820492050205120522053205420552056205720582059206020612062206320642065Average

Maximum Water Level

*Assumption here is 1981 and 2046 have the same initial condition (dam water level, discharge and volume of reservoir)Maximum Water Level

High Water Demand Season

High Water Demand Season

Snow & Glacier

Clean Glacier Debris covered glacier

Snow

Spatial distribution of snowfall

Fully Physical Model Upper Indus

Coupling with Climate Models

Seasonal variation of LAI by the Coupled Model (above) and MODIS (below)

River Discharge (obs. & simulated) LAI (obs. & simulated)

Sawada, Koike, et al. WRR (2014)

- Agricultural Drought Index -

Drought indices (SA index)Green:simulated annual peak LAI and Orange:nationwide crop production

The drought index calculated from the model-estimated annual peak of leaf area Index correlates well with the drought index from nationwide annual crop production.

Severe droughts (food shortage) in 1988-1989 and 1994-1995 are reported on FAO report [FAO, 2005]

R =0.89

Dro

ught

Sawada, Koike, et al. WRR (2014)

＋

Land surface model Dynamic Vegetation Model

AMSR-E(Aqua)

AMSR2(Shizuku)

Yang, Koike, et al. JMSJ (2007)

Electronic-MagneticWave

Sawada & Koike, JGR (2014)

Data AssimilationCoupled

18

GPM-Core

GCOM-W1

Sawada & Koike, JGR (2016)

Immediately after the flood Before the flood

Channel change in the middle reach of Akadani Debris flow deposition in the upstream of Akadani basin 19

Sediment Disaster Simulation: Debris Flow, Sediment Transport, Flood

20

Integrated Activity: Sediment Disaster Simulation in Northern Kyushu

Quantifying uncertainty

Climate models

Multi-model

ensemble (MME)

Down-scaling

Basin-scale

prediction of

quantity & quality

Water quantity

and quality prediction

flood

ordinary water

drought

ground water

Information

Storage

Treatment

Current facility, plan, management

Flood control system

Water allocation &

cost

Environment

Human life

Industry

HumanBehavior

EconomicBehavior

Drought Simulation

Flood SimulationIm

pact assessment

Filed survey

Early warning

Allocation policy

Land use

Adaptation options

Innovative technology- Flood control

- quality control

Decision m

aking

Monitoring evaluation

implem

entation

IntegratedObserved Data Sets

ProcessStudy

21

Scientific approach Engineering Approach Socio-economical approach

End to End Approach on Climate Change Adaptation

22

Flood duration= 1-2 days Flood duration= 3-4 days Flood duration= 5-6 days

Flood duration= 7 days Flood duration >7 days

0

20

40

60

80

100

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2

Perc

enta

ge o

f yie

ld lo

ss (%

)

Flood depth (m)

0

20

40

60

80

100

0 0.5 1 1.5 2

Perc

enta

ge o

f yie

ld lo

ss (%

)

Flood depth (m)

0

20

40

60

80

100

0 0.5 1 1.5 2

Perc

enta

ge o

f yie

ld lo

ss (%

)

Flood depth (m)

0

20

40

60

80

100

0 0.5 1 1.5 2

Perc

enta

ge o

f yie

ld lo

ss (%

)

Flood depth (m)

Vegetative Stage

Maturity Stage Ripening Stage Note: Green line and blue line are overlapped

Reproductive Stage

Flood Risk: Crop Damage Function

Developed based on flood damage matrix published by the Philippines Bureau of Statistics (2013) and considering height of rice plant

被害率

Flooded duration (day)

Life cycle Stage

Rice-Crops

37M Peso 1327M Peso 1952M Peso

Flood Damage Assessment in Pampanga River Basin

23

25-Year 50-Year 100-Year

Flood Contingency Scenario

24

Pampanga River

Labangan River

Angat River

Bag Bag River

Labangan River

Community Alert System using poles

Workshops for sharing results and discussing Barangay/Municipal Plan

Flood Contingency Planning

25Probability map of first flood inundation

Barangay Santa Lucia

Workshop at Municipality (2015.2)

Workshop at Barangays (2016.1)

Explanationof scenario

Discussionwith localresidents

Identification of necessaryaction for developing Barangay contingency plan(what they do? What they request)

Quantifying uncertainty

Climate models

Multi-model

ensemble (MME)

Down-scaling

Basin-scale

prediction of

quantity & quality

Water quantity

and quality prediction

flood

ordinary water

drought

ground water

Information

Storage

Treatment

Current facility, plan, management

Flood control system

Water allocation &

cost

Environment

Human life

Industry

HumanBehavior

EconomicBehavior

Drought Simulation

Flood SimulationIm

pact assessment

Filed survey

Early warning

Allocation policy

Land use

Adaptation options

Innovative technology- Flood control

- quality control

Decision m

aking

Monitoring evaluation

implem

entation

IntegratedObserved Data Sets

ProcessStudy

Scientific approach Engineering Approach Socio-economical approach

26

Volume Veracity Velocity Verity VisualizationData & Information

End to End Approach on Climate Change Adaptation

High Speed NetworkAnalysis Server

Extra-large volume data storage (25PB)

Base SystemICT Experts

Data Archive

Search / Download

Data ProcessingApplication Development

ICT Experts

Field Specialists R&D CommunityICT Experts

Field Specialists WaterDisaster

RiskReductionAgriculture

Urban

Economy

Biodiversity

Health Climate

Hydroelectric power

Social Implementation

Climate Change Adaptation

ASIAN Monsoon Year

International Contribution

DIAS/CEOSWater Portal

GEOSS/AWCI GEOSS/AfWCCI

Joint Research

S-8

CMIP5

GRENE-ei

DIAS-P

RECCA

Data Integration & Analysis System: Challenges to 5V

DIAS-ICHARM: Flood Information Sharing Support in Sri Lanka

In-situ rain gauge data (6 numbers)

Satellite precipitation data（GSMaP）

On-line Information provision on DIAS：In-situ rainfall, satellite rainfall, calibrated and forecast rainfall, inundation simulations

Implemented by EDITORIA and ICHARM on DIAS

ALOS © JAXA (2016)

Inundation analysis by using

RRI in DIAS

Simulation and forecasting of river

discharge, water level, inundation extent

Inundation analysis results

Concept of RRI model

Ensemble forecasting rainfall for the next 16 days (max)

Himawari-8 cloud images

Inundation map by satellite data (ALOS-2)

4 hr latency data (NRT)

Real time data (NOW)

Calibration

500mm

Bias-corrected Satellite Rainfall

Real-time Rain Gauge Data

Inundation

EnsembleFlood

Prediction

72hr11ensembleevery 24hr

Issued on May 24

Issued on May 25

Prediction

Water-Energy BudgetHydrological Model

(WEB-DHM)

River DischargeSoil MoistureGround WaterDam Storage

Crop Production

Satellite-based Land Data Assimilation

(CLVDAS)

Real Time Data Management System

DIAS Archives

NASA GLDAS Optimized LDAS ParametersAMSR2MODIS

GFDLAPCC

Hydromet Data Satellite Seasonal Forecast Global CLVDAS Outputs

Ensemble Rainfall Prediction

Ensemble Drought Prediction

Optimized RTM Parameters

NCDC Global Met

JMA Reanalysis

LDAS Reanalysis Statistics

River DischargeSoil Moisture, Ground Water

Dam StorageCrop Production

Rainfall Pattern 1Rainfall Pattern 2

Rainfall Pattern n

Prediction Accuracy EvaluationBias Correction

Weighting

Hydrometeorology-Agriculture Droughts Prediction System

Drought analysisWheat production

LAI anomaly from CLVDAS

2007 Morocco Drought

Morocco Algeria Tunisia

32

Quantifying uncertainty

Climate models

Multi-model

ensemble (MME)

Down-scaling

Basin-scale

prediction of

quantity & quality

Water quantity

and quality prediction

flood

ordinary water

drought

ground water

Information

Storage

Treatment

Current facility, plan, management

Flood control system

Water allocation &

cost

Environment

Human life

Industry

HumanBehavior

EconomicBehavior

Drought Simulation

Flood SimulationIm

pact assessment

Filed survey

Early warning

Allocation policy

Land use

Adaptation options

Innovative technology- Flood control

- quality control

Decision m

aking

Monitoring evaluation

implem

entation

IntegratedObserved Data Sets

ProcessStudy

Scientific approach Engineering Approach Socio-economical approach

33

Volume Veracity Velocity Verity VisualizationData & Information

Working Together Capacity BuildingPlatform

End to End Approach on Climate Change Adaptation

Additional slides for capacity building

Identification

Policy-making Community of Practice

1. Data Archiving

2. Model Development

3. Societal Benefit

Creation

Societal Change(land use , population)

Climate Change

Monitoring

Prediction

Tran

sdis

cipl

inar

y

Inte

rdis

cipl

inar

y

Cap

acity

Build

ing

Integrated Risk Assessment

future

present

past

International Cooperation

Hig

h-Le

vel P

ract

ition

erH

igh-

Leve

l Pol

icy-

Mak

er

Damage HazardSocio-Economic

35

Platform on Water and Disasters

Platform on Water and Disasters

Identification

Policy-making Community of Practice

1. Data Archiving

2. Model Development

3. Societal Benefit

Creation

Societal Change(land use , population)

Climate Change

Monitoring

Prediction

Tran

sdis

cipl

inar

y

Inte

rdis

cipl

inar

y

Integrated Risk Assessment

future

present

past

International Cooperation

Hig

h-Le

vel P

ract

ition

erH

igh-

Leve

l Pol

icy-

Mak

er

Damage HazardSocio-Economic

37

Cap

acity

Build

ing

in collaboration with National Graduate Research Institute for Policy Study (GRIPS)

MasterCourse

• Working at a national disaster managementorganization back home, he already had hadbasic knowledge and experience on socio-economic analysis and trans-boundary rivermanagement when he started the program. AtGRIPS and ICHARM, he improved his socio-economic understanding, and began to learn amethodology for assessing climate changeimpact on floods.

2016-2017 Master’s Course Mr. GAMA Samuel Joseph, Malawi

38

In response to his presentation, UNDP has provided financial support amounting to US$ 16 million for my office to improve on the EWS through the “*Scaling-up the Use of Modernized Climate Information and Early Warning Systems Project (M-CLIMES)”* under the Global Climate Fund.

50 Years Return Period

9.1%

• A Brazilian student was from the National Departmentof Civil Protection and Defense. While studying atICHARM, he worked on research aiming to apply theseven Global Targets, defined in the Sendai Framework,to disaster management in Brazil, proposing how to usethese targets for achieving local disaster risk reduction.

2016-2017 Master’s Course Mr. Mikosz Lucas, Brazil

Relationship between SF indicator B(No. of affected people per 100,000 population) and expected change of basin average precipitation under RCP8.5 scenario

Indicator B(Flood)

Indicator B(Drought)

39

Platform on Water and Disasters

Identification

Policy-making Community of Practice

1. Data Archiving

2. Model Development

3. Societal Benefit

Creation

Societal Change(land use , population)

Climate Change

Monitoring

Prediction

Tran

sdis

cipl

inar

y

Inte

rdis

cipl

inar

y

Integrated Risk Assessment

future

present

past

International Cooperation

Hig

h-Le

vel P

ract

ition

erH

igh-

Leve

l Pol

icy-

Mak

er

Damage HazardSocio-Economic

40

STIPS DMP

ICHARM

Cap

acity

Build

ing

in collaboration with National Graduate Research Institute for Policy Study (GRIPS)

ShortCourse

MasterCourse

PhDCourse

Quantifying uncertainty

Climate models

Multi-model

ensemble (MME)

Down-scaling

Basin-scale

prediction of

quantity & quality

Water quantity

and quality prediction

flood

ordinary water

drought

ground water

Information

Storage

Treatment

Current facility, plan, management

Flood control system

Water allocation &

cost

Environment

Human life

Industry

HumanBehavior

EconomicBehavior

Drought Simulation

Flood SimulationIm

pact assessment

Filed survey

Early warning

Allocation policy

Land use

Adaptation options

Innovative technology- Flood control

- quality control

Decision m

aking

Monitoring evaluation

implem

entation

IntegratedObserved Data Sets

ProcessStudy

Scientific approach Engineering Approach Socio-economical approach

41

End to End Approach on Climate Change Adaptation

Volume Veracity Velocity Verity VisualizationData & Information

Working Together Capacity BuildingPlatform

STIPS

ICHARM DMP

Quantifying uncertainty

Climate models

Multi-model

ensemble (MME)

Down-scaling

Basin-scale

prediction of

quantity & quality

Water quantity

and quality prediction

flood

ordinary water

drought

ground water

Information

Storage

Treatment

Current facility, plan, management

Flood control system

Water allocation &

cost

Environment

Human life

Industry

HumanBehavior

EconomicBehavior

Drought Simulation

Flood SimulationIm

pact assessment

Filed survey

Early warning

Allocation policy

Land use

Adaptation options

Innovative technology- Flood control

- quality control

Decision m

aking

Monitoring evaluation

implem

entation

IntegratedObserved Data Sets

ProcessStudy

Scientific approach Engineering Approach Socio-economical approach

42

Volume Veracity Velocity Verity VisualizationData & Information

Working Together Capacity BuildingPlatform

End to End Approach on Climate Change Adaptation

coarse resolution

fine resolution

statistical model

Dynamical Downscaing:Localized climate information is generated using high resolution regional climate models (RCMs), driven by low resolution global climate models (GCMs), or using a variable resolution global model in which the highest resolution is over an area of interest.

observed large scale climate

Transfer function(statistical model)

observed small scale climate

predicted large scale climate

predicted small scale climate

Statistical downscaling

Down-scaling & Bias Correction