Real Risk

Hazard data development

– A spatially accurate river network is

one of the most important data layers for any flood model. For

modelling purposes, only major rivers were considered. A lower

resolution river network (1:1 millon) was used to identify major

rivers and a high resolution (1:50,000) was used to ensure that

it was spatially accurate. RMSI developed semi-automated tech-

niques to handle spatial corrections with a quick turnaround.

A 25-metre lateral resolution digital elevation model (DEM) was

used for the higher end GIS-based flood risk model. The vertical

resolution of the DEM was 1 metre. This DEM was hydrologically

corrected using semi-automatic techniques developed by RMSI,

to make it compatible with established GIS-based flood model-

ling. These corrections included sink filling, reconciliation of the

DEM with respect to the river network, removing artifacts etc.

Model simulations were carried out using the Federal

Emergency Management Agency approved hydraulic model

(HEC-RAS). For each river segment, water surface profiles were

simulated for flows ranging from two-year to 10,000-year return

periods. The computational procedure was based on a solu-

tion of one-dimensional continuity, energy and Manning’s

equations using the standard step method. Flood extents and

depths at 25-metre grid resolution were derived for each

stochastic event.

Flood depths were aggregated to various resolutions depend-

ing on the urban concentration. Flood hazard was calibrated for

recent events. Hazards included off-floodplain flooding from small

streams, sheetflow and drainage overflow.

Vulnerability functions

– The vulnerability functions in the

Belgium river flood model were based on cost modelling

methods that consider replacement and reinstatement costs for

each component of loss. The model was calibrated and vali-

dated using historical loss information and expert knowledge.

The vulnerability functions produce mean damage ratios for

buildings as well as contents coverage for a wide range of occu-

pancies and construction types. Coverage for residential

alternative living expense and commercial business interrup-

tion was also included. While most of the damage from

catastrophic floods occurs from inundation by large rivers,

significant losses also occur due to sheetflow, the flooding of

small streams, or backup of drainage systems. The vulnerabil-

ity model reflected these different flood types, with specific

relationships for river inundation and off-floodplain losses. The

hazard database and damage curves were then fed into the

client’s software to carry out financial loss computations.

The Belgium river flood model provides high-resolution capa-

bility for pricing and underwriting policies, and the effective

management of company-specific flood aggregates and expo-

sure.

Case study 2: Vulnerability and adaptation to drought –

economic impact scenarios

Drought sets off a vicious cycle of socio-economic impact begin-

ning with crop yield failure, unemployment, erosion of assets,

decrease in income, worsening of living conditions, poor nutrition

and subsequently, decreased risk absorptive capacity.

The state of Andhra Pradesh in India has historically been most

severely affected by drought. The failure of monsoons has had a

fatal affect on the state’s sizeable agriculture sector and on the

large share of the population dependent on agriculture for its

livelihood. The state has many ongoing programmes for drought

and watershed management. The World Bank wanted to comple-

ment the efforts of the state government by assessing the

economic and fiscal implications of drought, based on potential

climate change and risk mitigation scenarios.

The objectives of this study included developing a robust

analytical framework for simulating the long-term impact of

drought at the micro (drought-prone areas) and macro (state)

levels; conducting a quantitative probabilistic risk assessment of

the impact of drought under different scenarios; and assisting the

state government in the development of a forward-looking and

anticipatory strategy for adapting to frequent drought events and

conditions of water deficit.

In addition to macroeconomic and drought management

scenarios, the development of the modelling framework was

also aimed at assessing the possible increase in frequency and

severity of droughts that may occur due to human-induced

climate change.

The methodology for undertaking this study included devel-

oping a probabilistic drought risk assessment model that included

hazard, vulnerability and economic modules.

The hazard module included stochastic events simulating the

characteristics of historical events. A probabilistic drought risk

model in the form of a weather generator was used to analyse

and quantify the impact of potential future drought in the state

and to compute direct losses including probable maximum and

average annual losses.

A vulnerability module helped quantify the damage caused to

each crop due to weather hazards. An agro-meteorological model

was then used to analyse the impact of drought on crops. The

analysis included daily time step weather, rainfall distribution

and intensity, as well as deficiencies in surface and sub-surface

water supplies and soil moisture.

[

] 143

Hazard Module

Simulated weather generator

Indirect monetary losses

Vulnerability Module

Crop yield model

Planting area model

Indirect Loss Module

Input-output model

Macroecenomic model

Direct Loss Module

Translation of production losses into

monetary losses

Stochastic normal and

drought events

Production losses

Direct monetary losses

Historical weather

Crop soil

management

Commodity prices

Macroecenomic

data

Probabilistic drought risk assessment model including hazard,

vulnerability and economic modules

Source: RMSI