Case Pakistan
Muhammad Atiq Ur Rehman Tariq
Risk-based planning and optimization of flood
management measures in developing countries:
Case Pakistan
ter verkrijging van de graad van doctor
aan de Technische Universiteit Delft,
op gezag van de Rector Magnificus prof. ir. K.C.A.M. Luyben,
voorzitter van het College voor Promoties,
in het openbaar te verdedigen
op maandag 19 december 2011 om 15:00 uur
Muhammad Atiq Ur Rehman Tariq
M.Sc. in Engineering Hydrology
Centre of Excellence in Water Resources Engineering, University of
Engineering and Technology Lahore, Pakistan
geboren te Multan, Pakistan
Dit proefschrift is goedgekeurd door de promotor:
Prof. dr. ir. N. C. van de Giesen
Dr. ir. O. A. C. Hoes
Samenstelling promotiecommissie:
Rector Magnificus
Prof. dr. ir. N. C. van de Giesen,
Technische Universiteit Delft, promotor
Dr. ir. O. A. C. Hoes,
Technische Universiteit Delft, copromotor
Prof. dr. Ir. Pieter van der Zaag,
UNESCO-IHE Delft/ Technische Universiteit Delft
Prof. ir. Han Vrijling,
Technische Universiteit, Delft
Prof. dr. ir. Bart Schultz,
UNESCO-IHE Delft/ Technische Universiteit Delft
Prof. dr. ir. Arjen Hoekstra,
Universiteit Twente, Enschede
Dr. ir. P.H.A.J.M. van Gelder,
Technische Universiteit Delft
Prof. dr. ir. J.B. van Lier
Technische Universiteit Delft, reservelid
Keywords: risk-based flood management, developing countries, expected annual damages, optimal risk point, structural
measures, flood zoning, flood insurance
Copyright © 2011 by Muhammad Atiq Ur Rehman Tariq
Contact: [email protected]
Cover designed by: Dr. Van Tuan NGUYEN
ISBN: 978-90-6562-284-6
© All rights reserved. No part of the material, protected by this copyright notice, may be reproduced or utilized in any
form or by any means, electronic or mechanical, including photocopying, recording or by any information storage and
retrieval system, without prior written permission from the author.
Printed in the Netherlands.
‘Munir Akhtar’
̶ may her soul rest in the Heaven (Amin)
Although circumstances did not allow her to finish her own education, she succeeded to
motivate me, her son, to complete mine…
A number of persons, organizations, and groups have helped to shape my work as it is
presented in this thesis. I would thank Allah Almighty for his blessings that enabled me to
produce this research work.
When I chose TU Delft for my PhD, I was planning to do research filled with obscure
terminology and complicated formulae. Questions of Prof. Nick van de Giesen and Dr.
Olivier Hoes forced me to think first on the fundamentals of flood management practices
that, as a result, changed my topic, motivations, and the targets of research. I am indeed
thankful for their critical and innovative questions and their support.
Scholarship for this research was awarded by HEC Pakistan under the cherished
motivation of then chairperson Prof. Dr. Ata Ur Rehman. I would cordially thank him, who
turned the education opportunities towards the deserving students under open merit
I would like to extend my sincere thanks to Adam J. Pel, Shahzad Akhtar, and Sandra
Junier for sparing their precious time to review chapters of this thesis. As well as, Martine
Rutten, who translated the Summary and Qianqian Zhou, who provided innovative ideas
employing adaptation possibilities, thank you both. Many thanks are due to Yasir Abbas
for providing necessary data. I am also thankful to all of my colleagues for their support to
facilitate my research and helping me in many ways. I wish to stay in contact with all of
Back in Pakistan, my father Al Haaj Shafi Muhammad Bhatti who remained always
waiting for my successful return. I cannot forget the support of my seniors, Brig. (R) Ali
Mansoob Raza and Engr. Ayub Tabassam, who made all efforts for me to proceed with my
PhD. My friend Wasif Ali always motivated me to complete my PhD as early as possible. I
always pray for them from the core of my heart.
My wife Maria, my daughter Eshaal Areej Fatima and my sisters always remained my
sound support during my stay in the Netherlands. I would thank my brothers Shafiq Ur
Rehman and Laiq Ur Rehman as well as my in-laws Mr. Majeed Baloch & family for all
their help in every aspect of my life.
Muhammad Atiq Ur Rehman Tariq
Problem statement
Among all the natural hazards, floods claim most lives and highest financial losses. About
95-97% of all deaths and a significant part of economic losses caused by floods occur in
developing countries. Despite spending considerable resources, flood management
arrangements in developing countries are still unable to deliver satisfactory results.
Limited resources, lack of research, and absence of proper planning are the main
constraints towards the optimization of flood management in developing countries. Flood
management plans are often accomplished with the financial and technical help of
developed countries following their design methodologies and safety standards. Such
plans are often not suitable for local conditions. Flood safety standards for developed
countries may not produce the most efficient results when applied in developing
The subject of this thesis is flood management in developing countries, focusing on fluvial
floods, with special focus on Pakistan. To address this issue, the development of a
methodology that can appropriately consider the socio-economic and technical
constraints of a country is crucial. This thesis intends to provide guidance for flood
managers and land-use planners that develop a floodplain. Their general concerns can be
described as: How much should land-use planners curtail developments and how much
should the river be regulated to minimize flood risks? For a satisfactory solution of these
concerns, a risk-based methodology is introduced and outcomes are compared with
locally used and worldwide well-established methods. In other words, the objective of
this thesis is to develop a method ‘to maximize advantages by minimizing flood
deductions in a floodplain, using the available resources’.
Flood management in Pakistan
Flood management in Pakistan is a resource-demanding and complex issue. A complete
understanding of the problem and the existing setup is important. The nature of the
problem of flooding varies extremely over the country depending on the hydrological,
topographical, and demographical conditions. Considerable resources have been
deployed over time since the country came into being in 1947. Despite deploying major
financial resources and institutional support, flood management is still not appropriately
optimized at the national level. The recent devastating flood in 2010 raised concerns over
the reliability of existing flood management arrangements.
Risk-based flood management
Like most countries, Pakistan also uses probability-based flood safety standards (using a
50-years return period) for flood protection structures. To design flood protection
measures according to probability-based standards, only flood characteristics need to be
considered. Moreover, these standards do not provide guidance for the design of nonstructural measures. A risk-based flood management approach provides the logical
grounds for the selection and design of flood measures, whether structural or nonstructural. It incorporates fairness and uniformity, and provides a firm basis to flood
management practices. Risk calculations should not ignore the advantages of floodplain
usage for productive purposes, like residential, industrial, and agricultural use. Riskbased analysis provides a systematic methodology to optimize flood management.
Components of flood risk
To estimate and minimize the risk of flooding at any location within the floodplain, it is
important to know the basic components of risk and understand their role in inducing
risk. Although risk mechanics have been defined differently in literature, this research
proposes alternative, more appropriate, concept that considers all independent
components of risk while avoiding redundancy. In this research, risk is defined as the
result of the interaction between ‘hazard’ and ‘vulnerability’. While a hazard can be
characterized by its intensity and probability, vulnerability depends on ‘susceptibility’
and ‘exposure’. The understanding of risk mechanics is important to select appropriate
flood management measures with optimized design.
Cost benefit analysis
Cost benefit analysis (CBA) is a requisite for risk-based assessment. Multi-objective or
multi-criteria analysis (MCA) may also be incorporated as a part of CBA for the
optimization purposes. Careful valuation of economic, social, and environmental assets is
indispensable for the precision of CBA. These analyses are often used to determine the
feasibility of projects nowadays. Benefit to cost ratio (BC Ratio), Present Value (PV),
Internal Rate of Return (IRR), and Economic Rent (ER) are economic efficiency indicators
commonly used for the evaluation of projects. These indicators help to best utilize
investments, but do not explicitly correlate the targeted flood management arrangements
to the optimal arrangements.
Expected annual damages
The proposed risk-based assessment method helps decision-makers to envisage the
spatial and temporal distribution of risk over the floodplain. Direct and indirect losses
due to floods are related to the occurrence of floods, whereas investments in measures
consist of the initial capital investments and the maintenance costs throughout the useful
life of the measures. The ‘Expected Annual Damages’ (EAD) combine both the flood losses
and investments in measures and spread (average) these over the lifetime of measures.
Damage curves and EAD distribution maps provide a detailed picture of risk distribution.
Optimal risk point (ORP)
Nowadays, flood management aims at deriving maximum net-benefits from a floodplain.
Starting from this premise, a method to maximize the net benefit is developed. Generally,
flood losses decrease as investments on flood measures increase. Marginal costs for
measures to reduce losses are initially low. When efforts are made to further decrease
flood losses, the ratio between marginal benefits and marginal costs reduces. The point
where marginal losses are equal to marginal benefits is known as the ‘optimal state’.
Different measures can achieve different optimal states. If different measures are
considered to reduce flood losses and to increase land-use benefits, then a number of
optimal states can be obtained. The optimal state with the lowest risk is called the
‘Optimal Risk Point’ (ORP). Combined direct, indirect, and induced losses due to floods
are at their minimum while maximum benefits from the floodplain can be obtained at this
point. Once the ORP is determined, economic indicators can provide guidance for the
feasibility of a project.
Case studies
Two case study areas were selected. The first study area is the 90km long reach of the
Chenab River from Marala Headworks to Qadarabad Headworks in Pakistan. It is densely
populated, with valuable economic activities, and it has an almost flat terrain. The second
study area is located on the Swan River within Islamabad City (Capital of Pakistan) where
land is very expensive and developers have an interest in developing every inch of land.
Risk assessment and flood management optimization of these areas provided valuable
information and suggested solutions for land-use developers and floodplain inhabitants.
The following measures were analyzed:
Risk-based design of dike crest levels
Risk-based flood zoning
Risk-based flood insurance
In the study area of the Chenab River (from Marala to Qadarabad), flood risk was found to
be extremely high where the General Trunk (GT) Road crosses the river connecting Gujrat
and Wazirabad cities. The reduced width of the floodway causes the inundation of urban
areas due to the backwater effect. At present, dikes are provided downstream of the GT
Road bridge as the river slope is gradual and the area is almost flat. In those rural areas
that experience frequent flooding, people adapt their land-use pattern to cope with this
situation by reducing exposure and/or susceptibility. In the study area of the Swan River,
risk had increased severely due to recently built residential areas. A number of private
dikes divert hazard impacts to the other side of the River. These dikes are designed either
with arbitrary design standards or with probability-based flood standards to protect the
inner areas. Dikes and dredging were found not to be suitable measures here, due to the
high relief of the area and the steep slope of the riverbed. Flood zoning was demonstrated
to be the most effective measure in this study area.
Development of effective and efficient flood management addressing the socio-economic
conditions of developing countries is important. The concept of zero flood risk as a result
of total flood protection is unrealistic in a floodplain. The acceptability of risk is a function
of social, economic, and environmental concerns and may vary largely at personal and
national levels. Arbitrary or probabilistic standards do not consider this variation and
will not lead to an effective and efficient design of measures. Optimization of flood
management requires adjustments to reduce the hazard as well as the vulnerability. The
ORP-concept may help in developing risk-based standards for flood management.
Flood management must be an integral part of land-use development, in this way
reducing risk by providing both land-use modifications and flood protection measures.
Risk assessment must be accompanied by an intense understanding of social, economic,
and technical aspects of a country and floodplain. The national risk management policy
should consider the local context and address the issues accordingly. The personal choice
of accepting flood risk may be covered by introducing complementary insurance or tax.
Providing information indicating the initial risk and the residual risk after taking
measures is extremely important. Therefore, the standard method practiced in a country
should assure the information generated portrays the entire context. In addition, more
reliable and agreed upon approaches must be developed to incorporate social and
environmental factors into risk calculations.
Muhammad Atiq Ur Rehman Tariq
Overstromingen eisen de meeste levens en veroorzaken de grootste economische schade
van alle natuurrampen. Ongeveer 95-97% van de overstromingsslachtoffers valt in
ontwikkelingslanden en ook een belangrijk deel van economische schade treedt hier op.
Overstromingsbeheer-maatregelen leveren nog steeds geen bevredigend resultaat. De
optimalisatie van het overstromingsbeheer wordt beperkt door gebrek aan middelen,
onderzoek en goede planning. Overstromingsbeheerplannen worden meestal opgesteld
en uitgevoerd met financiële en technische hulp van ontwikkelde landen, op basis van
hun ontwerpmethodieken en veiligheidsnormen. Dergelijke plannen passen vaak
onvoldoende bij lokale omstandigheden. Hoogwaterveiligheidsnormen ontwikkeld voor
ontwikkelde landen leiden vaak niet tot de meest efficiënte resultaten bij gebruik in
Het onderwerp van dit proefschrift is het overstromingsbeheer in ontwikkelingslanden,
met een focus op rivieren. Een methodiek die adequaat rekening houdt met de socioeconomische en technische beperkingen van een land is hierbij van groot belang. Tevens
probeert dit proefschrift richtlijnen te bieden voor waterbeheerders en planologen die
overstroombare gebieden langs de rivieren inrichten en beheren. Hun dilemma kan
grofweg als volgt worden omschreven: in hoeverre moeten zij enerzijds socioeconomische ontwikkelingen in die gebieden beperken en anderzijds de rivier beteugelen
om overstromingsrisico’s te minimaliseren. Voor goed onderbouwde afwegingen
hierover werd een op risico gebaseerde methodiek geïntroduceerd en werden de
resultaten vergeleken met de lokale praktijk en wereldwijd gevestigde methoden. Het
doel van dit proefschrift kan in het kort omschreven worden als 'de voordelen van
overstroombare gebieden maximaliseren en tegelijkertijd de kosten minimaliseren,
gebruik makend van beschikbare middelen’.
Waterveiligheid in Pakistan
Overstromingsbeheer in Pakistan is complex en vraagt de inzet van veel middelen. Een
volledig begrip van het probleem en de huidige aanpak is belangrijk. De aard van het
overstromingsprobleem varieert sterk over het land, afhankelijk van de hydrologische,
bodemkundige, en demografische omstandigheden. Sinds het land in 1947 onafhankelijk
werd, zijn aanzienlijke investeringen gedaan, maar ondanks die inzet en de institutionele
ondersteuning, is het beheer van overstromingen op nationaal niveau nog steeds niet
geoptimaliseerd. De verwoestende overstroming in 2010 voedde de bezorgdheid over de
betrouwbaarheid van de bestaande overstromingsbeheersmaatregelen.
Overstromingsbeheer op basis van risico
Net als in de meeste landen van de wereld worden beschermingsmaatregelen in Pakistan
ontworpen op basis van hoogwaterveiligheidsnormen die de kans op overstromingen
uitdrukken (bijvoorbeeld een herhalingstijd van 50-jaar). Bij het ontwerp van een
maatregel op basis van deze normen, worden alleen de overstromingkarakteristieken
meegenomen. Deze normen geven daarnaast geen richtlijnen voor het ontwerp van nietstructurele maatregelen. Een op risico gebaseerde aanpak kan de selectie en het ontwerp
van alle overstromingbeheermaatregelen ondersteunen, zowel structurele als nietstructurele. Een dergelijk aanpak is in hoge mate betrouwbaar en uniform, en biedt solide
grondslagen voor overstromingsbeheer. De risicoberekening moeten niet voorbijgaan
aan de voordelen van het gebruik van overstroombaar gebied voor productieve
doeleinden, zoals wonen industrie en landbouw. Op risico gebaseerde analyses bieden
een systematische methode om overstromingsbeheer te optimaliseren.
Componenten van overstromingsrisico
Om het overstromingsrisico op elke overstroombare locatie in te schatten en te
minimaliseren, is het belangrijk om de basiscomponenten van de risico's te kennen en te
begrijpen wat hun rol is in het induceren van risico's. Hoewel in de literatuur ook andere
risicomodellen zijn gedefinieerd, wordt in dit proefschrift een voor dit onderzoek
geschikter concept voorgesteld dat alle onafhankelijke componenten van het risico
beschouwt terwijl het redundantie vermijdt. In dit onderzoek is risico gedefinieerd als
resultaat van de interactie tussen 'gevaar' en 'kwetsbaarheid'. Een gevaar kan worden
gekenmerkt door ‘intensiteit’ en ‘waarschijnlijkheid’; kwetsbaarheid door ‘gevoeligheid’
en ‘blootstelling’. Begrip van de mechanismen die risico’s veroorzaken is belangrijk om
geschikte overstromingsbeheermaatregelen te selecteren en het ontwerp hiervan te
Een kosten-batenanalyse (KBA) is nodig voor een op risico-gebaseerde beoordeling. Een
multi-criteria analyse (MCA) kan ook worden opgenomen als onderdeel van KBA's voor
optimalisatiedoeleinden. Een zorgvuldige bepaling van de economische, sociale en
ecologische waarde is essentieel voor de precisie van de KBA. Deze analyses worden
tegenwoordig veel gebruikt om de haalbaarheid van projecten te beoordelen. De ratio van
kosten en baten, de contante waarde, de interne opbrengstvoet en de rentabiliteit zijn
economische efficiëntie-indicatoren die gebruikt worden voor de evaluatie van projecten.
Deze indicatoren helpen om investeringen zo goed mogelijk in te zetten, maar leiden niet
Verwachte jaarlijkse schade
De op risico gebaseerde beoordelingsmethode helpt om de ruimtelijke en temporele
verdeling van hoogwaterrisico’s in overstroombare gebieden inzichtelijk te maken voor
besluitvormers. Directe en indirecte schade als gevolg van overstromingen treden
incidenteel op, terwijl voor het treffen van maatregelen zowel eenmalige initiële
investeringen als investeringen in onderhoud gedurende de levensduur nodig zijn. Het
concept de ‘verwachte jaarlijkse kosten’ (VJK) combineert deze twee soorten kosten en
spreidt ze uit over de looptijd van de maatregelen. Schadecurven en VJKverspreidingskaarten geven een gedetailleerd beeld van de spreiding van de risico's.
Optimaal risico punt (ORP)
Het huidige overstromingsbeheer heeft als doel overstroombare gebieden maximaal te
benutten. Dit is het uitgangspunt voor het ontwikkelen van een methode waarbij de baten
worden gemaximaliseerd. In het algemeen neemt de hoogwaterschade af naarmate de
investeringen in maatregelen toenemen. De marginale kosten om verliezen te beperken
zijn in het begin laag. Wanneer er meer inspanningen worden gedaan om verdere daling
van hoogwaterschade te bewerkstelligen, wordt het verschil tussen de marginale baten
en marginale kosten kleiner. Het punt waarbij de marginale verliezen gelijk zijn aan
marginale baten wordt de 'optimale toestand' genoemd. Verschillende maatregelen
kunnen leiden tot verschillende optimale toestanden. Als verschillende maatregelen
worden beschouwd om overstromingsverliezen te verminderen en de opbrengst uit
landgebruik te verhogen, zijn meerdere optimale toestanden mogelijk. De optimale
toestand met het laagste risico staat bekend als het 'optimale risico punt’ (ORP). Directe,
indirecte en geïnduceerde verliezen als gevolg van overstromingen zijn op dit punt
minimaal, terwijl de voordelen van het gebruik van deze gebieden maximaal zijn. Na het
bepalen van het ORP kunnen economische indicatoren toegepast worden om de
haalbaarheid van de projecten te beoordelen en een prioriteitsvolgorde te bepalen.
Case studies
In twee gebieden zijn case-studies uitgevoerd. Het eerste studiegebied is de rivier de
Chenab over een lengte van 80 kilometer tussen Marala Headworks en Qadarabad
Headworks in Pakistan. Dit is een dichtbevolkt gebied met veel economische activiteit en
een bijna vlak terrein. Het tweede studiegebied is gelegen aan de Swan River in
Islamabad (hoofdstad van Pakistan), waar het land erg duur is en ontwikkelaars elke
centimeter grond willen ontwikkelen. Risicobeoordeling en optimalisatie van
overstromingsbeheer in deze gebieden biedt waardevolle informatie en oplossingen voor
planologen en inwoners van overstroombare gebieden. De volgende maatregelen zijn
• Risico-gebaseerd ontwerp van dijken
• Risicozonering
• Risico-gebaseerd overstromingsverzekeringsstelsel
In het studiegebied van de Chenab River (van Marala tot Qadarabad), werd vastgesteld
dat het overstromingsrisico extreem hoog is op de plek waar de General Trunk (GT)
Road, die de steden Gujrat en Wazirabad verbindt, de rivier kruist. De verminderde
breedte van het rivierbed op dit punt veroorzaakt opstuwing en daardoor worden
stedelijke gebieden voor de brug overstroomt. In plattelandsgebieden die frequent
overstromen, passen de mensen hun landgebruik zodanig aan dat de kwetsbaarheid en
de blootstelling wordt verminderd.
In het studiegebied van de Swan River bleek dat het risico sterk is toegenomen als gevolg
van recent gebouwde woonwijken. Door enkele private dijken wordt het gevaar
verplaatst naar de andere zijde van de rivier. Deze dijken zijn ontworpen op basis van
willekeurige ontwerpnormen of op kansberekening gebaseerde hoogwaterstanden om de
achterliggende gebieden te beschermen. Dijken aanleggen en baggeren zijn hier geen
geschikte maatregelen vanwege het grote reliëf van het gebied en de steile helling van de
rivierbedding. Overstromingsrisicozonering bleek de meest effectieve maatregel in dit
De ontwikkeling van een effectieve en efficiënte aanpak van overstromingsbeheer die
past bij de sociaal-economische omstandigheden van de ontwikkelingslanden is
belangrijk. Het concept van verwaarloosbare overstromingsrisico’s en volledige
bescherming tegen overstromingen is niet realistisch. De aanvaardbaarheid van risico is
een functie van sociale, economische en omgevingsfactoren, en kan sterk variëren op
persoonlijk en nationaal niveau. Willekeurige of probabilistische normen houden geen
rekening met deze variatie en leiden niet tot een effectieve en efficiënte vormgeving van
maatregelen. Optimalisatie van het overstromingsbeheer vraagt om aanpassingen gericht
op zowel het gevaar als de kwetsbaarheid. Het ORP concept kan helpen bij de
ontwikkeling van op risico gebaseerde normen voor het overstromingsbeheer.
Het overstromingsbeheer moet een integraal onderdeel van de ruimtelijke ordening zijn
om risico's te verminderen door middel van aanpassingen in landgebruik zowel als
beschermingsmaatregelen. De risicobeoordeling moet gebaseerd zijn op goed begrip van
sociale, economische en technische aspecten van een land en het overstroombare gebied.
Nationaal risicobeheer beleid moet passen bij de lokale context en de lokale problemen.
De persoonlijke keuze van het aanvaarden van overstromingsrisico kan gedekt worden
door de invoering van een aanvullende verzekering. Het leveren van goede informatie is
uiterst belangrijk, de standaard in een land gebruikte methode moet informatie leveren
die een compleet beeld van de situatie geeft. Betrouwbaardere en maatschappelijk meer
gedragen benaderingen moeten worden ontwikkeld om sociale en ecologische factoren te
integreren in risicoberekeningen.
Translated by
M.M. Rutten and Sandra Junier
Acknowledgements ....................................................................................................................... vii
Summary ............................................................................................................................................. ix
Samenvatting.................................................................................................................................. xiii
Introduction ............................................................................................................................. 1
Flood management practices ................................................................................................... 2
Problem statement ....................................................................................................................... 3
Major factors ................................................................................................................ 3
Possible way forward ............................................................................................... 4
Objectives ......................................................................................................................................... 4
Study areas ....................................................................................................................................... 5
The Swan River ........................................................................................................... 6
The Chenab River ....................................................................................................... 7
Structure of thesis ......................................................................................................................... 7
Introduction..................................................................................................................................... 9
Floods due to cyclones in Makran coastal area ............................................................. 16
Floods and flood management in Pakistan .................................................................... 9
Floods ................................................................................................................................................. 1
Fluvial floods in the Indus basin .......................................................................................... 11
Flash floods in the Indus and Kharan basins .................................................................. 14
Flood management arrangements ...................................................................................... 17
Flood management measures ............................................................................................... 17
Structural measures ............................................................................................... 18
Non-structural measures ..................................................................................... 20
Legal framework ......................................................................................................................... 22
Institutional arrangements .................................................................................................... 23
Hazard managing institutes ................................................................................ 23
Crisis managing institutes ................................................................................... 24
Analysis and discussion ........................................................................................................... 26
2.10 Discussion...................................................................................................................................... 28
Risk-based assessment of flood management............................................................ 30
Flood impacts ............................................................................................................................... 30
Components of flood management practices ................................................................. 31
Measures ..................................................................................................................... 33
Plans/ projects ......................................................................................................... 37
Enhancement in flood management................................................................ 38
Identification of problem ..................................................................................... 39
Holistic considerations ......................................................................................... 39
Integrated approach .............................................................................................. 39
Incorporation of diverse measures ................................................................. 40
Balanced approach ................................................................................................. 40
Uniformity and fairness ........................................................................................ 41
Long-term considerations ................................................................................... 41
Risk-based assessment ............................................................................................................ 41
Risk ................................................................................................................................ 42
Hazard .......................................................................................................................... 43
Vulnerability.............................................................................................................. 45
Calculation of direct damages ............................................................................ 45
Risk-based flood management practices .......................................................................... 47
Risk-based classification of measures ............................................................ 47
Risk-based classification of approaches ........................................................ 48
Risk-based explanation of plans/ projects ................................................... 50
Development of decision-support system ....................................................................... 50
Assessment/ approval criteria .......................................................................... 34
Guidelines for flood management approach ................................................................... 38
Approaches ................................................................................................................ 34
Guiding principle for decision-support system .......................................... 50
Cost benefit analysis (CBA) ................................................................................. 51
Economic efficiency indicators .......................................................................... 52
Discounting procedures ....................................................................................... 55
Optimal risk point (ORP) ..................................................................................... 56
Discussion...................................................................................................................................... 58
River behavior (Hazard parameters)................................................................................. 60
Flood management assessment framework ............................................................... 60
Flood frequency analysis ..................................................................................... 62
Rainfall-runoff modeling for Swan study area ............................................ 64
Flood simulations .................................................................................................... 66
Societal response (Vulnerability parameters) ............................................................... 66
Land-use practice (Exposure) ............................................................................ 67
Scale of the study ..................................................................................................... 69
Tangible losses ......................................................................................................... 70
Intangible losses ...................................................................................................... 71
Expected annual damages (EAD)...................................................................... 72
The unit area (Yardstick) approach ................................................................ 73
Software tools ........................................................................................................... 73
Uncertainty and risk analysis ............................................................................. 74
Design optimization .................................................................................................................. 75
Discussion...................................................................................................................................... 76
Hazard adjustments ............................................................................................................ 77
Flood response (Susceptibility) ........................................................................ 67
Risk assessment .......................................................................................................................... 67
Design of structural measures .............................................................................................. 78
Residual risk .............................................................................................................. 78
Socioeconomic development and flood management ............................. 79
Induced risk ............................................................................................................... 79
River dynamics ......................................................................................................... 79
Risk-based design of dikes and dredging: Case studies ............................................. 80
Dike-crest level design (the Swan River case) ............................................ 81
Dike-crest level design (the Chenab River case) ........................................ 84
Risk-based river dredging ................................................................................... 87
Discussion...................................................................................................................................... 89
Flood mapping ............................................................................................................................. 92
Vulnerability adjustments ................................................................................................ 91
Types of flood maps .................................................................................................................. 92
Flood mapping practices ......................................................................................................... 94
Flood mapping in the USA ................................................................................... 94
Flood mapping in Germany ................................................................................. 95
Flood mapping in the UK...................................................................................... 95
Flood mapping in the Netherlands .................................................................. 95
Flood mapping in France ..................................................................................... 96
Flood mapping in Pakistan .................................................................................. 96
Flood zoning ................................................................................................................................. 96
Resistance and reluctance against flood zoning......................................... 97
Risk-based flood zoning ....................................................................................... 98
Flood zoning for the settlements ................................................................... 106
Case studies .................................................................................................................................. 99
Concluding remarks ............................................................................................................... 109
Flood insurance in practice ................................................................................................. 111
Flood zoning for agricultural land.................................................................... 99
Vulnerability indirect-adjustments ............................................................................. 111
Risk-based flood insurance ................................................................................................. 112
Recovery from flood............................................................................................ 113
Flood risk awareness .......................................................................................... 113
Reduction in damages ........................................................................................ 114
Advantages of risk-based insurance................................................................................ 114
Discussion................................................................................................................................... 120
Case study ................................................................................................................................... 115
Results and conclusions........................................................................................................ 115
Conclusions and recommendations............................................................................. 121
Discussion and conclusions ................................................................................................ 121
Limitations ................................................................................................................................. 122
Recommendations .................................................................................................................. 123
References ...................................................................................................................................... 125
Annex A ........................................................................................................................................... 136
Annex B ........................................................................................................................................... 141
Annex C ............................................................................................................................................ 146
Annex D ........................................................................................................................................... 149
Annex E ............................................................................................................................................ 153
Annex F ............................................................................................................................................ 161
List of Figures
Figure 1-1: Map of Pakistan indicating the Swan River at General Trunk Road, Rawalpindi
and Chenab River floodplain from Marala Headworks to Qadarabad Headworks.
Figure 1-2: Temporal distribution of precipitation at the Swan River basin
Figure 1-3: Outline of this thesis
Figure 2-1: Map of Pakistan rendering the important hydrological and geopolitical features
Figure 2-2: Historical flood flows in main rivers (1921-2010), Source: NESPak13
Figure 2-3: Locations of telemetric gages, HF radios, weather radars, and river structures
Figure 2-4: Flood losses details at national level against severe flooding years, Sources: FFC
annual report 2008 and National Disaster Management Authority
Figure 2-5: Area flooded and crop area flooded at district level for major rivers, Sources:
District governments and Punjab Irrigation department
Figure 3-1: Schematic concept of Measures, Approaches, Design Criteria, and Projects
Figure 3-2: Risk mechanics explained in terms of hazard, vulnerability, probability,
intensity, exposure, susceptibility, and consequences
Figure 3-3: Comparative risk analysis of different strategies and calculation of optimal risk
against different land-use schemes in floodplain
Figure 3-4: Combined deductions and minimum required net-benefit of land-use
Figure 3-5: Relationship between increasing costs on measures and reducing flood damages
Figure 3-6: Comparative risk analysis of different strategies and calculation of optimal risk
point against a land-use scheme in floodplain
Figure 3-7: Flood management costs, benefits, BC ratio, NPV, and ORP concepts58
Figure 4-1: Proposed schematic layout of risk-based assessment
Figure 4-2: Flood frequency analysis for both study areas: the Chenab River at Marala
Headworks using Log Pearson Type-III distribution, [MM method], the Swan River
flows at Dhok Pathan and annual maximum rainfall at Islamabad Airport using
Generalized Extreme value Distribution, [ML method]
Figure 4-3: Schematic layout of HEC-HMS rainfall-runoff model of the Swan River catchment
Figure 4-4: Land-use at 1km resolution on left (Source: (Cheema and Bastiaanssen, 2010))
and right side figure represents the hydro-topographical features of the Swan
River study area catchment at 25m resolution.
Figure 4-5: Chenab River floodplain showing the digitized land-use details. 68
Figure 4-6: The Swan River floodplain showing the developed areas by different developers.
Figure 4-7: Depth-damage function used for case studies
Figure 4-8: Comparative risk analysis of different strategies and calculation of optimal risk
point against a land-use scheme in floodplain (Reproduced).
Figure 5-1: Damage curves of existing, baseline case, and continuous dikes of different
heights show the trends in flood losses for the Swan River.
Figure 5-2: Damage curves of existing, baseline case, and fragmental dikes of different
heights show the trends in flood losses for the Swan River.
Figure 5-3: EAD distribution maps of existing, baseline, fragmental dike, and continuous
dike (both at 4m heights) for the Swan River.
Figure 5-4: Damage curves of existing, baseline case, and continuous dikes of different
heights show the trends in flood losses against floods for the Chenab River. 85
Figure 5-5: EAD distribution maps of existing case show the spatial distribution of EAD over
the Chenab River reach
Figure 5-6: EAD distribution maps of baseline case show the spatial distribution of EAD over
the Chenab River reach
Figure 5-7: EAD distribution map of proposed dike showing the spatial distribution of EAD
against 5m high proposed dikes over the Chenab River reach
Figure 5-8: Damages curves of existing, baseline case, and continuous dredging of different
depths show the trend in flood losses for the Swan River
Figure 5-9: Damages curves of existing, baseline case, and continuous dredging of different
depths show the trend in flood losses for the Chenab River
Figure 5-10: EAD distribution map of 5m deep continuous dredging over the Swan River
Figure 5-11: EAD distribution map of proposed 5m deep dredging over the Chenab River
Figure 5-12: Relationships between the increasing efforts to curtail flood losses and losses
due to floods and costs spent on measures to determine ORP for the Swan River.
Figure 5-13: Relationships between the increasing efforts to curtail flood losses and losses
due to floods and costs spent on measures to determine ORP for Chenab River.
Figure 6-1: A virtual flooding scenario showing weighted flood spreads based on their
probabilities of occurrence in study area, from Marala to Qadarabad, the Chenab
Figure 6-2: Relationship between flood depth and residual net-profit for agricultural land.
Figure 6-3: Maps show the distribution of expected annual net-profit of agricultural land for
the Chenab River reach against the existing conditions, before and after zoning.
Figure 6-4: Maps show the distribution of expected annual net-profit of agricultural land for
the Chenab River reach against the baseline, before and after zoning.102
Figure 6-5: Maps show the distribution of expected annual net-profit of agricultural land for
the Chenab River reach against 6m dikes, before and after zoning.103
Figure 6-6: Loss reduction economic effect of risk-based zoning for the Chenab River area.
Figure 6-7: Impacts of agricultural land-use zoning in reducing expected annual damages for
the Chenab River area.
Figure 6-8: Area selected in the 5km buffer of river centerline to evaluate the increase in
net-profit for the Chenab River case.
Figure 6-9: Increase in agricultural land-use net-profit for 5km buffer area around the
Chenab River.
Figure 6-10: Impacts of dwellings land-use zoning in reducing expected annual damages for
the Swan River area
Figure 6-12: Floodplain area selected within 300m of the river centerline for the Swan River
case study.
Figure 6-13: Increase in residential land-use net-profit for 300m buffer area around the
Swan River.
Figure 6-11: EAD distribution maps of baseline case, existing, and 4m fragmental dikes
proposed zoning for the Swan River reach
Figure 7-1: Spatial distribution of flood risk considering all probable floods.116
Figure 7-2: Spatial distribution of flood losses due to 1% exceedance flood. 116
Figure 7-3: The calculated damage against 1% probability design flood on all stations in
study area.
Figure 7-4: The calculated risk calculated damages considering all probable floods on all
stations in study area.
Figure 7-5: Comparison of estimated average damages based on 1% probability flood with
calculated damages considering all probable floods.
Figure 7-6: Comparison of estimated cumulative damages based on 1% probability flood
with calculated damages considering all probable floods.
Figure 7-7: Schematic cross section of the Chenab River elaborating the general slope of
ground and shifting of risk due to dikes which are mostly provided on left-bank
side of the River.
List of Tables
Table 2-1: Main features of the Indus River System
Table 3-1: Classification of flood impacts (river floods)
Table 2-2: Details of embankments and spurs at the provincial and national levels
Table 3-2: Daily life hazards with negligible risk
Table 3-3: Flood management measures and target risk parameters
Table 4-1: Details of methods selected while modeling rainfall-runoff using HEC-HMS64
Table 4-2: Land-use classification used in our case studies for calculation of direst tangible
Table 4-3: Popular software tools, their compatibility to GIS and their country of origin
1 Introduction
Among all the natural catastrophes, floods have claimed most lives and have caused the
highest economic losses than any other single natural hazard in the world. Based on data
of floods and all other natural hazards used in long-term analysis (1988-1997)1, the
following facts about floods shows the severity of problem (Loster, 1999):
Floods count 1/3 of all natural catastrophes (largest counts)
Floods claim 10% of all the fatalities due to natural catastrophes
Floods cause about 1/3 of the overall economic loss
Floods have caused approximately $250 billion loss worldwide within last 10
Floods occur throughout the world with large variations in their impacts. There are areas,
for example, the Nile, the Mekong, and the lower Indus floodplains, where regular flooding
is a source of agricultural activities. However, with growing population and industrial
activities in floodplains, floods are not accepted anymore in majority of floodplains. As a
result, flood management and flood control measures are introduced in many places to
control floods and their negative consequences (Duivendijk, 1999).
Actually, floodplains have been the sources of food, livelihood, and economic activities.
Consequently, these floodplains have always been centers for human settlements,
especially in early ages of human civilization. Since then, society is taking advantages
associated with floodplains while trying to avoid flood losses. Due to high densities of
masses and economic activities in floodplains, the impacts of floods are felt severely. The
Netherlands and Bangladesh are examples of floodplains that exhibit high population
densities (APFM, 2009a).
Although these analyses (by Munich-RE) are quite old, these provide comprehensive comparisons.
Some of newly conducted analyses support the same trends e.g., EM-DAT http://www.emdat.be/.
Losses due to natural disasters have greater impacts on developing countries
comparatively. The largest casualties due to floods occur in developing countries and Asia
(Douben, 2006). About 95-97% of all deaths and major part of economic losses caused by
disasters occur in developing countries (Tung, 2002; WMO, 2003; UN-WATER, 2005; Kelly
and Garvin, 2007). Naturally, such damages are higher proportions of national income in
developing countries (Dilley et al., 2005). Flooding events do not affect only developing
nations, but often affect and devastate the economically advanced and industrialized
nations. Recent tsunami in Japan in March 2011 is an example. In the USA, 90% of all
natural disasters involve flooding (Collins and Simpson, 2007) and costs approaching $6
billion annually (ASFPM, 2003, 2004).
Due to demography, life-style, and climatic trends, flood damages are likely to become
more frequent, prevalent and serious as time passes (Borrows and Bruin, 2006). These
higher flood losses demand an effective system for flood management and present
growing trends in flood losses emphasize a revision of existing strategies on priority. An
understanding of floods, their impacts, and flooding mechanics is necessary to propose
improvements in flood management.
Flood management practices
Intensive use of floodplain triggered ever-growing flood losses. Structural measures are
considered as conventional and effective measures. The first interaction with river can be
the one when human settlements started maintaining a minimum safe distance from the
rivers. Hence, the flood zoning (in an informal way) can be considered as the first step
towards flood management (Cuny, 1991). The use of structural measures for flood
protection dates back to the civilization of the Nile and the Romans. Structural measures
remained the only option for a long time, but it was realized then that embankments are
not suitable for all situations.
As mentioned, first action of man against floods was non-structural consisting of living at
higher and safer areas, it was in the late 1950’s, when non-structural measures were
introduced as engineered measures. In addition, it was realized that the cost of protecting
an area liable to flooding with structural measures is sometimes higher than for nonstructural measures (Duivendijk, 1999; Tucci, 2007). Non-structural measures, such as
warning systems, floodplain zoning, flood proofing, and flood insurance programs started
to receive more attention. Further improvements were brought into the field of flood
management with recent concepts of flood management. Integrated, sustainable, riskbased, holistic, and balanced concepts of flood management as well as the concepts of
floodplain conservation are examples. Now these concepts are conjointly applied in the
field of flood management. Sustainable and effective flood management requires an
integration of policies, plans, projects, strategies, and measures (Petry, 2002).
Problem statement
Unfortunately, establishment of effective and optimized strategies are still important
issues in developing countries (Petry, 2002). Limited resources, lack of research, and
absence of proper planning are often considered main restraints towards the optimization
of flood management in developing countries. Flood management plans are sometimes
accomplished with the financial and technical help of developed countries along with their
design methodologies and safety standards. Such plans do not suit to the local conditions
appropriately. High safety standards (500-1000 years return period) proposed by donor
countries for Dhaka after floods in 1988 are a good example that failed to provide desired
results during floods in 1998 (Stalenberg and Vrijling, 2009). Flood management
standards devised for developed countries do not efficiently produce desired results in
developing countries. In this perspective, development of a methodology that considers
the socio-economic and technical constraints of developing countries is crucial.
Nevertheless, the flood losses in developing countries are expected to increase in future
due to their increasing vulnerabilities and absence of flood management strategies. In
recent decades, developing countries have increased their vulnerabilities against floods
due to their rapid development (ADPC, 2005). If no serious step is taken now, the risk will
continue to increase in the future. Unfortunately, with the existing setups, it seems that
“disasters will continue to take people, communities, and governments by surprise in
developing countries” (White et al., 2001).
Major factors
Before handling the flood problem in developing countries, understanding the flooding
characteristics is very important. The flooding situation is worsening due to typical
unplanned land-use development style and ad-hoc flood management practices. Most of
new settlements are taking place with ‘unplanned and organic’ pattern (APFM, 2009a).
Economic activities are usually concentrated in big cities. The poor populations, which
earn their livelihood from big cities, have to acquire these outskirts of metropolises that
are particularly at risk (WMO, 2003). Mostly, there is no insurance coverage to such
settlements or people do not have the capacity to buy one. In addition, the ratio of
economic loss to the total economy tends to be higher in future (WMO, 2003).
As mentioned earlier, most of flood management programs are accomplished with the
financial help of international donors. The fact is that governments and NGOs put the
interests of their national companies and members ahead of those of the recipients who
live in floodplains (Green et al., 2000). These lopsided arrangements cannot assure a
successful solution, instead, will develop mismatched adjustments.
Another important factor is that developed nations of the world produce the majority of
greenhouse gases. The impacts will be more severe on developing countries who have
larger vulnerable populations, national economies dependent on agricultural production
and are not fully equipped to deal with extreme flooding (Pelling et al., 2004). As those
losses are uninsured, affected communities take a long time to recover. Sometimes, floods
revisit these communities before they are fully recovered from the previous episode.
Possible way forward
The developing countries must progress in the field of disaster mitigation and
vulnerability reduction (White et al., 2001), because, without such reforms, developing
countries will have little chance of generating higher economic growth. One hurdle in the
way of sustainable flood management setup in developing countries is their dependency
on post disaster aid, which can be considered as a problem of moral hazard (Benson and
Clay, 2004). Developing countries should learn from the experiences of the developed
countries rather than to replicate their policies and strategies (Green et al., 2000). Due to
increasing trends in flood losses, even developed countries must revise their flood
management practices. There are no ‘off the shelf’ management solutions that are
invariably more appropriate than the others. As every climate and hydrological situation
(rain, river, and sea proximity etc) is different from others, every floodplain needs to be
treated individually. A general assessment method applicable to all floodplains, however,
can be established. In addition, some common characteristics for a successful strategy can
be defined there to evaluate the suitability of an approach. A genuine and logical approach
must be proposed, rather than simply transposing the approaches of developed countries.
The motive behind this research is to improve the flood management in developing
countries that are characterized by their limited resources, dense population, and
unplanned urbanization. The initial idea was to take advantage from the experiences of
developed countries. Although, developed countries that experience comparatively less
damages, are not an exception from flood damages. It was planned to see how these
experiences could be used for developing countries. Due to the scope of work, this thesis
focuses flood management in general, and fluvial floods in developing countries, in
particular. It will help in developing a balanced, affordable, and effective approach at the
national level.
This thesis intends to assist flood managers and land-use planners in the process of
floodplain development and maintenance. The general concerns of both sides can be
described as such: How much the developments are to be restrained and how much river
should be trained to ensure a harmonious coexistence with floods? For the satisfactory
solution of these concerns, risk-based methodology is introduced and outcomes are
compared. In other words, the purpose of this thesis is to facilitate maximizing the
advantages and minimize damages in floodplain within societal capacity. The focus is set
to establish a criterion that can help to evaluate the suitability of flood measures for
developing countries. The objective of research can be defined as “to develop a standard
flood management assessment approach that aims at optimizing the land-use net-benefits
by reducing flood deductions in a floodplain suitable to socio-economic conditions of a
country and the local society”. This assessment may help to choose an appropriate flood
management strategy suitable to any country whether it is a developed or developing
country. That criterion may also help in choosing the most efficient measure or
combination of measures to produce maximum benefits of floodplain.
Flood management is a technical problem, which has to address many aspects of legal,
institutional, communication, emergency management, environmental, monitoring, and
land development issues as well. It has to address all the components of flood risks. It
must allow freedom to consider diverse measures. Assessment of measures, in our case,
will be done within a probabilistic framework through risk-based criteria. A comparison
with existing and presently practiced methods will also be performed. As mentioned, the
flood management experiences in developed countries will also be utilized to achieve our
Study areas
For our case studies, two floodplains in Pakistan are selected. The Chenab River and the
Swan River were selected. The river reach from Marala Headworks to Qadarabad was
selected on The Chenab River whereas the reach between Islamabad Highway and GT
Road was selected on the Swan River. Figure 1-1 shows the locations of study areas.
Figure 1-1: Map of Pakistan indicating the Swan River at General Trunk Road, Rawalpindi and
Chenab River floodplain from Marala Headworks to Qadarabad Headworks.
The selection of the study areas is based on three reasons
Pakistan represents a good example of a developing country where floods are
common and efforts are made to establish an appropriate flood management
strategy at the national level;
Data was easily available and accessible due to the cooperation of government
and private departments;
The research is funded by the Higher Education Commission of Pakistan.
Although there was no such demand by the Commission, Pakistani rivers were
selected as goodwill to honor its financial support;
The Author is familiar with the river system and floodplain characteristics of the
study areas.
The Swan River
The Swan River flows through Islamabad city. Due to intense pressure of extending
population areas, developers are developing its floodplain. In the absence of a standard
evaluations procedure, this development is going unchecked. In some areas of floodplain,
development of residential areas might be extremely risky. Few dikes are presented in
area that shift flooding to other side of river causing increase in risk across the river.
These dikes may shift risk upstream or downstream as well. These are designed either
with arbitrary design standards or against a certain flood design. People interested in
buying property in protected areas never know that these developments might be safe
against some design flood but not safe against all floods. Whereas, the standards adopted
are also highly debatable.
Figure 1-2: Temporal distribution of precipitation at the Swan River basin
The Chenab River
The Chenab River is selected for this study as being the major source of fluvial floods in
Pakistan. Having no suitable site for major storage reservoir in Pakistan, the Chenab River
flows as an uncontrolled river. Whilst, Baghlayar and Salal hydropower dams in India (run
of the river projects) have little effect on flood alleviation because of their limited storage
capacities and mismated reservoir operation practices. Due to continuous deforestation in
upstream hilly catchment areas, even higher floods are expected in future. The Chenab
River is 1,240 km long and drains a basin of 67,500 km2 excluding of its major tributaries
Jhelum, Ravi, and Sutlej and joins the Indus River (NESPak, 2008). As mentioned, 90 km
long river reach between Marala Headworks and Qadarabad Headworks has been
selected. The selected reach is situated after the River just crosses international boundary
and flood-warning system for this initial reach is not very effective.
Structure of thesis
This dissertation provides a conceptual approach for the assessment of flood management
strategies based on risk process and its spatial distribution, equipped with interactive
flood management. Figure 1-3 describes the schematic layout of this thesis.
Chapter 2 describes flood characteristics and the implemented flood management system
in Pakistan. Basic concept of risk and risk-based flood management is proposed in
Chapter 3. Chapter 4 provides the complete framework for the assessment of any
measure, plan, or strategy. From Chapter 5 to Chapter 7, are the case studies and
implementation demonstrations of proposed assessment method. Chapter 5 deals with
the impacts of hazard adjustment. Chapter 6 presents the impacts of reducing community
vulnerability. Whereas, Chapter 7 gives a brief description of impacts of flood insurance
and indicates its importance to achieve desired vulnerability adjustments. Chapter 8
contains the main conclusions of this work and recommendations for future work.
Figure 1-3: Outline of this thesis
2 Floods and flood management in Pakistan
In August 2010, Pakistan suffered one of the most severe floods in its history. Floods are
the most frequently occurring and damaging natural hazards in the country. Of all
population who are affected by natural hazards, 90% are subjected to flooding (Haider,
2006). In the recent flooding, almost 1800 persons died and financial damages were in
range of tens of billions US dollars. According to available official statistics, about 8,000
people lost their lives and economic losses amounted to approximately $10 billion
between independence in 1947 and the 2010 flooding (Baig, 2008). These estimates are
carried out at the local administration level and uncertainty in these values is unknown.
Although no major flood had occurred since 1995, the devastating flooding in 2010
demonstrated the continuous presence of flood risks.
The nature of flooding varies according to geography. Fluvial floods in the Indus plain
prove most devastating, as the terrain is flat, densely populated, and economically
developed. High discharges in rivers coming from upstream are the main cause of floods.
Hill torrents (flash flooding) are the second most destructive type of flood. Hill torrents
threaten large areas of the country (Figure 2-1) and claim human lives most frequently.
Floods due to cyclones and intensive localized rain are dominant at other locations.
Exceptionally high floods have also occurred due to the breaching of some of the small
dams, e.g. the Shadi Kor dam in Pasni, which breached on February 11, 2005, washing
away more than 135 people (IFRC, 2005; Javed and Baig, 2005). The hydrology of floods is
linked to weather and climate as well as to physiographic features (Shah and Gabriel,
2002). A brief overview of related geographical features is provided to interpret the
flooding characteristics. The country can be divided into three physiographical regions
(Framji and Mahajan, 1969):
Mountains in the north and north-west 241,647 km2
Plateau of Baluchistan in the south-west 242,683 km2
Indus River plains 311,766 km2
This Chapter is reproduced from the author’s article published in Journal of Physics and Chemistry
of the Earth with DOI 10.1016/j.pce.2011.08.014
Figure 2-1: Map of Pakistan rendering the important hydrological and geopolitical features
The spatial variability of rainfall throughout the country is high. Of the total area, 59.3%
can be classified as rangeland, which receives less than 200 mm annual rainfall (Umrani,
2001; ISDR, 2005). In the north of the country, the Himalaya Range receives annual
rainfall between 760 mm and 1270 mm (ISDR, 2005) and contributes almost 72% of the
mean annual flow in the Indus River System (WWF, 2010). These rainfall data are based
on the national meteorological network. The spatial distribution of stations over the
country is not uniform. Stations in developed areas and meteorologically important
locations generally comply with World Meteorological Organization (WMO) standards.
Southern Punjab, Baluchistan, and northern Sindh receive the lowest amounts of rain.
Rainfall increases again towards the coast. Three types of weather systems influence the
precipitation in catchments, which produce floods in Pakistan. These weather systems are
Monsoon depressions originating from the Bay of Bengal (the most important
Westerly waves coming from the Mediterranean Sea (Winter rains)
Seasonal lows from the Arabian Sea (Cyclones)
The country has four distinct climate seasons. April, May, and June are extremely hot and
dry months. July, August, and September are hot and humid with intense heat and heavy
but scattered rainfall (monsoon). The cool and dry period starts in October and continues
through November. December, January, and February are the coldest months of the year.
Hydrologically, the country can be divided into three major units: Indus basin, Kharan
basin, and Makran coastal drainage area. Flooding characteristics of these basins vary
greatly and require in-depth understanding.
Fluvial floods in the Indus basin
The total watershed area of the Indus is 944,000 km2, 60% of which lies in Pakistan (MoE,
2003). The Indus, with its major tributaries Jhelum, Chenab, Sutlej, and Ravi, has an
average annual flow of 175 km3/yr. Table 2-1 presents a brief overview of the major
rivers in the Indus Basin.
Seasonally, flows fluctuate from 3,000 m3/s to 34,000 m3/s (FFC, 2009). Annual river
flows at rim stations (first gauging station after a river enters into Pakistan) fluctuate
between 120 km3/yr and 230 km3/yr (MoWP, 2002b). Rainfall in the Indus Basin occurs
during the monsoon and cold weather seasons, but severe floods only occur in the
monsoon season. High flows are experienced in the summer due to the increased rate of
snowmelts and monsoon rainfalls. About 82% of the annual water flows during the
summer months (MoWP, 2002c). In this period, heavy rainfall in the upper catchments
located across the border in Kashmir (Indian) often causes floods. Sometimes heavy
showers occur in areas just within Pakistan. As a consequence, the rivers expand into
their entire floodplains. The flooding behavior of the major rivers differs according to
catchment characteristics and the types of installed river training facilities. In low
elevation catchments (Sutlej, Ravi, and Jhelum), maximum snowmelt occurs in April
through June and does not coincide with the monsoon rains (July through September). In
high altitude catchments (Indus and Chenab), snowmelt contributes significantly to flood
flows. Maximum snowmelt in the Indus and Chenab basins is experienced in July and
floods of high magnitude are generated due to monsoon rainfalls. The flood peaks of the
different rivers usually do not coincide. However, when they do coincide, widespread
flooding occurs.
Table 2-1: Main features of the Indus River System
Basin Area (km2)
Avg. Annual Flow
Dams in India
Discharge to
Length (km)
Dams in Pakistan
No. of Barrages
Sources: WAPDA annual report 2000, FPSP-II/C report 2008, and FFC annual report 2008
‫ ٭‬Himachal Pradesh, India
ᵔ Pong Dam on the Beas River (Major tributary of the Sutlej River)
Floods in the Indus and Jhelum basins are largely controlled by the Tarbela and Mangla
dams. There is almost no control (in Pakistan) over the Chenab, Ravi, and Sutlej rivers,
which results in flooding problems in the monsoon season. The Chenab has historically
given rise to the most devastating floods because of the lack of any controlling structures
and large flows induced by the combination of rain and snowmelt. India owns the
exclusive water rights of the Ravi and Sutlej rivers under the Indus Water Treaty (1960).
Because of that, there is practically very little flow in these rivers (Haq and Nasir, 2003).
Average annual flows observed at the rim stations are about 3.15 km3/yr in the Ravi River
and 0.02 km3/yr in the Sutlej River (Mir et al., 2010). Floods of higher intensity are
observed on the Ravi River after the Treaty. According to annual peak flows data at the
Balloki Barrage, of the seven most severe floods on the Ravi River (1922-2004), six floods
occurred after effectuation of this treaty in 1973 (Figure 2-2). The decreasing width of
these rivers and vanishing flows encourage encroachments for residential and industrial
purposes, but an episode of severe flood may wipe out these developments.
Figure 2-2: Historical flood flows in main rivers (1921-2010), Source: NESPak
In the upper and mid reaches of the Indus Basin, it is generally the tributaries like the
Jhelum and the Chenab rivers that cause flooding rather than the Indus River itself. Since
these rivers are also snow-fed, an early monsoon may combine with peak snowmelt
runoff to exacerbate flooding. Generally, heavy rainfalls are limited to the Chenab, Jhelum,
Ravi, and Sutlej River catchments. Occasionally, low atmospheric pressure crosses further
north into the Indus River catchment like in the recent case of flooding. Intense rainfall
produced exceptionally high flood peaks, which resulted in flash flooding in North West
Frontier Province (NWFP, now Khyber-Pakhtunkhwa) and fluvial flooding in Punjab and
Sindh provinces. Fluvial flooding caused losses by inundating large agricultural and
residential areas, by damaging lifelines and powerhouses, and by eroding land along the
The nature of fluvial floods in the upper Indus Plain differs from that of the lower Indus
Plain. In the upper Indus, the bed level is lower than the adjoining lands. When a flood
occurs, floodwater spilling over the riverbanks generally returns to the rivers in the upper
part of the Indus Basin. However, in the lower part of the Indus in Sindh province, where
the riverbed is higher than the floodplain (suspended river), spills do not return to the
river. This lack of return flow extends the duration of inundation, resulting in larger
damages. Although flood protection by embankments has been provided along almost the
entire length in Sindh province and at many locations in the upper areas, bund breaches
can still occur (Haq and Nasir, 2003; Khan, 2007b; FFC, 2009). Such breaches often cause
greater damage than would have occurred without dikes because of their unexpected
nature and intensification of land-use following the provision of flood protection.
Flash floods in the Indus and Kharan basins
Flash floods typically hit the hilly areas of NWFP, Baluchistan, Kashmir, and Punjab.
Kashmir and NWFP receive high average annual rainfall, whereas the steep and barren
terrain of Baluchistan and Dera Ghazi Khan (D.G. Khan) watersheds typically produce
severe flash floods, causing damage to infrastructure, settlements, and loss of human and
animal lives. Flash flooding in the Indus Basin, is confined to the tributaries of the Indus,
Jhelum, and Chenab rivers. Most areas in NWFP, Kashmir, and Baluchistan and some areas
in Punjab endure flash floods. Flash floods are relatively lethal, e.g., more than 230 people
died due to flash floods in the Pothohar Plateau (Islamabad, Rawalpindi, and NWFP areas)
in 2001 (IFRC, 2002). According to flood loss data of the Federal Flood Commission (FFC),
about 60% of the casualties were reported in NWFP during the 2010 flood due to flash
flooding. Consolidated economic loss and casualty data has not been compiled nationwide
and very little flood discharge data for hill torrents is available. It is extremely difficult to
measure such peak flows with conventional methods due to their short duration and their
Floods in the NWFP are mainly hill torrents due to steep bed slopes, which greatly
increase flood velocity and severely erode the banks. To save the areas from erosion,
spurs have been constructed by the provincial government with the funds provided by the
federal government. Fluvial floods in NWFP are limited to Nowshera and some parts of
Charsadda, Peshawar (by the Kabul River), and Dera Ismail Khan (by the Indus River). In
the rest of NWFP, flash flooding is a common disaster along with landslides and torrential
rains (PMD, 2009). Some dikes have been provided for flood diversion or abatement as
well as to minimize the effects of torrential rains and consequent floods. Other severe
flash flooding occurs in Dera Ismail Khan along the Indus. These hill torrents have an
average annual flow of about 1 km3/yr (MoWP, 2002a). A battery of spurs has been
constructed on the right bank of the Indus River (FFC, 2009). Large numbers of spurs and
a few embankments have been constructed along the Swat, Kurrum, and Kabul rivers and
their tributaries.
The area of the Pothohar plateau (in north Punjab) often experiences flash flooding.
Islamabad and Rawalpindi have endured flash floods from the Nullah Lai, which nearly
flows through the centers of both cities. The low-lying areas in Rawalpindi along the
Nullah Lai are even affected by small floods. Extreme floods in Nullah Lai were observed
in 1981, 1988, 1997, and 2001 (Kamal, 2004). The hill torrents generated in Suleiman
Ranges (Baluchistan and Afghanistan) hit the districts of D.G. Khan, Layyah, and Rajanpur
in Punjab province. As the catchment area that generates torrents is quite far away from
the above-mentioned districts, sometimes, weather conditions in the catchment area and
affected areas are very different. In such cases, these torrents appear without any weather
symptom or warning sign. D.G. Khan hill torrents have an average annual flow of 1 km3/yr
(MoWP, 2002a). These floods have destroyed bridges, settlements, and agricultural land
along riverbanks and have deposited huge amounts of debris into the rivers.
All of Baluchistan Province, with its barren and steep land, is subject to hill torrents. The
Nari, Kaha, and Gaj rivers are part of the Indus drainage system located in the
northeastern edges of the province. Contrary to the rest of Baluchistan, the Kachi area is
highly fertile and needs floods for irrigation (Jarrige, 1997).
Kharan Basin (within Pakistan) covers an area of 121,860 km2 and includes part of the
Kharan Desert and Pishin Basin in west Baluchistan. Average annual rainfall throughout
the desert is less than 100 mm (Khosa, 2000) and average inland drainage is about 1
km3/yr (Shah and Gabriel, 2002; UNITAR, 2004). The flow regime in the rivers is typified
by spring runoff and occasional flash floods caused by Westerly waves during the winter
months. The riverbeds are dry for most of the year. Intense flash floods do occur but are
infrequent. Some bunds have been constructed to serve as flood diversion or abatement
measures. During a severe flood, most of the embankments and floodwalls constructed to
protect orchards or abadies (residential areas) are washed away. As flash floods of high
intensity that disturb the settlements are quite rare, people are not prepared for disaster
responses, which results in relatively large destruction and losses.
Pishin Lora Basin is a major river basin in Baluchistan (16,928 km2 with 10 sub-basins)
spread over five districts with a total population of about 1.2 million (ADB, 2008). As this
basin covers the area with Baluchistan’s main economic activities and high population
concentration, the disturbance due to floods is high.
Floods due to cyclones in Makran coastal area
The coastal area of Pakistan stretches over a length of 1,046 km between 62oE and 68oE
(Rehman and Bhattarai, 2005). Makran in the south of Baluchistan is a semi-desert coastal
strip with an area of 123,025 km2 and a length of 750 km along the Arabian Sea (Shah and
Gabriel, 2002). The region is sparsely populated, with much of the population being
concentrated in small ports and fishing villages. Away from the coast, the narrow coastal
plain rises very rapidly into several mountain ranges. The entire length of the coastline is
subjected to tropical cyclones. The Makran Coastal Basin includes the Dasht, Hingol, and
Porali rivers, which discharge individually into the Arabian Sea (MoWP, 2002a) with an
average annual flow of 3.5 km3/yr. The climate is dry with very little rainfall and can be
classified as arid with warm summers and mild winters. The monsoon rainfall increases
with the increase in longitude along the coastline, whereas winter rainfall decreases with
the increase in longitude. The average annual rainfall is approximately 150 mm or even
less along the Makran Coast.
Floods in coastal areas are associated with cyclones and high tides. The Makran Coastal
Areas have occasionally been hit by severe cyclones. Cyclones generated in the Arabian
Sea produce torrential rains throughout the region. One cyclone is expected per year in
the Arabian Sea. About 75% of these cyclones end up at the Omani coast on the western
Arabian Sea and the remaining 25% curve clockwise and cross the coast near the Rann of
Kutch (MoE, 2003). No severe tidal floods have been recorded so far. The coastal areas of
Sindh are the most vulnerable and most exposed to cyclones. The period from 1971-2001
saw 14 cyclones (ISDR, 2005). One severe cyclone in 1997 affected Makran (Gawadar and
Kech) and then crossed into the Kharan Basin up to the Chaghai and Dalbadin districts.
The Nihang and Kech rivers caused widespread flooding in a region approximating 8,000
km2 (PMD, 2009). The floods due to heavy showers of two consecutive cyclones caused
tremendous damage. Cyclone Gonu struck the coast on June 4, 2007 and inflicted damage
in the Sur Bandar area of Gawadar (Khan, 2007a). Cyclone Yemyin on June 26, 2007 is
among the worst recorded so far. It affected 2.5 million people and made 250,000
homeless (UNESCO, 2007). The cyclone hit the catchment area of the Mirani Dam (Dasht
River). Substantial rainfall occurred during the storm, causing serious flooding in the
Dasht River. The Pakistan Meteorological Department data showed rainfall of 172 mm
over the storm period (two days) at Turbat Airport in Baluchistan. The rainfall event was
the highest rainfall recorded in the last 90 years (NDMA, 2007). The storm moved from
east to west, moving from the Kech River’s catchment to the Nihing River’s catchment, the
two main tributaries of Dasht River. As mentioned earlier, intense cyclones do not occur
often, but they can cause large-scale damage and cyclone Yemyin was one such example.
This cyclone caused flash flooding in various districts of the Baluchistan and Sindh
Flood management arrangements
After independence, devastating floods occurred in 1950, 1956, and 1957. Due to limited
resources and institutional arrangements, no comprehensive flood management plan was
initiated at the national level. Until 1976, flood protection and management was the sole
responsibility of provincial governments. This changed after the annihilating floods of
1973, which claimed 474 lives and caused damages of 160 billion Pakistani Rupees (PKR)
(approximately $2 billion) (FFC, 2009). A unified countrywide approach was initiated to
manage the flood problem. As a result, a long-term principal plan was prepared in 1978 at
the national level. The present flood management arrangements can be discussed under
three aspects:
Flood management measures
Legislative framework
Institutional setup
Flood management measures
The flood management measures in Pakistan are mainly comprised of flood protection
embankments, spurs, studs, and advanced flood-forecasting techniques. Various flood
protection structures were built by the provincial governments to solve local flood
problems (Baig, 2008). Since the establishment of FFC in 1977, flood management has
been practiced according to an integrated approach at the national level. A long-term
National Flood Protection Plan (NFPP) was prepared in 1978. The NFPP contained phased
implementation in the form of sub-plans known as the ‘ten-years National Flood
Protection Plans’ (NFPPs). An estimated expenditure of over PKR 17.8 billion
(approximately $220 million) has been spent on flood works, rescue and relief not
included, under different programs since 1977 (FFC, 2009). A number of flood protection
works have been completed and some are still in the implementation phase. The
provinces receive financial and technical support provided by the FFC to address the
flooding problem.
So far, three NFPPs have been executed covering periods from 1978-1987 (NFPP-I), 19881997 (NFPP-II), and 1998-2007 (NFPP-III). Under NFPP-I, 350 flood protection schemes
(individual structure repaired or constructed) were implemented at a cost of PKR 1.73
billion (approximately $22 million) (Shaikh, 2008). NFPP-II was carried out under two
sub-projects, namely, the Normal Annual Development Plan (NADP) and the Flood
Protection Sector Project-I (FPSP-I). Under FPSP-I, 170 schemes (costing PKR 2,541
million, approximately $32 million) have been completed under the NADP and 257
schemes (costing PKR 4,860 million, approximately $61 million) have been executed.
Three sub-projects were carried out under NFPP-III (1998-2007). 101 schemes (costing
PKR 4,165 million, approximately $52 million) under FPSP-II, 362 schemes (costing PKR
3,415 million, approximately $43 million) under the NADP and development of a flood
forecasting and warning system for Lai Nullah in Islamabad/ Rawalpindi (PKR 348
million, approximately $4.5 million) have been completed for this plan (FFC, 2009). These
plans have been financed by the government and some donor agencies. The execution of
the flood protection works is the responsibility of the provincial agencies, while decisionmaking and control of funds lie with the federal government. The approving authority for
each single sub-project is also the federal government. About PKR 17.8 billion
(unadjusted, approximately $222 million) has been spent on flood management measures
since 1977 and about PKR 30 billion (approximately $375 million) is planned for NFPP-IV
(2008-2017) (Shaikh, 2008; FFC, 2009). Financial resources, employed in rescue, relief,
and rehabilitation process are used in addition to the above-mentioned expenditures.
According to the planning and approval criteria of the FFC, new flood projects are
executed under two categories: either need-based measures to address local flood
problems or integrated measures under the NFPP. Priorities are given to those measures,
which serve areas of high economic losses, human suffering, and socially and
economically vulnerable groups. Since the NFPP plans are mostly financed through loans
from the Asian Development Bank, the measures are not sanctioned unless they have an
economic internal rate of return (EIRR) of at least 12% (FFC, 2009) in compliance with
bank criteria. The EIRR of a project can be defined as the average annual effective
compounded return rate of investments. EIRR serves to enable a direct comparison of
investments and benefits, which typically have a different temporal distribution. EIRR is
very common indicator in a cost-benefit analysis. Protection standards adopted in
Pakistan are 50-years for flood protection structures and 100-years for vital river training
structures and bridges (Halcrow et al., 2001). The planning and approval criteria are the
same throughout the country, but there are different practices locally in design,
construction, and maintenance of bunds, studs, and spurs.
Structural measures
Numerous efforts have been made in the past to train rivers and protect the adjoining
areas from river erosion and flood damages, but large-scale variations in river discharge
and sediment concentrations have led to eroding river plains. Flood management plans
initiated at the government level have relied heavily on the provision of structural
measures for flood containment. Structural measures are employed on a large-scale and
include construction of embankments, spurs, dikes, gabion walls, floodwalls, dispersions,
diversion structures, delay action dams, bypass-structures, and channelization of
floodwaters. River training has mainly been executed with the help of embankments and
spurs. Embankments are constructed wherever over-bank flooding is the major problem
and spurs are constructed to counter land erosion and regulate the river’s main course.
About 6,719 km of embankments have been constructed along major rivers and their
tributaries. In addition, more than 1,375 spurs have been constructed to protect these
embankments (FFC, 2009). Details of embankments and spurs at provincial and national
levels are provided in Table 2-2. Economical and efficient measures have been
implemented based on their suitability for local conditions. For the most part, earthen
dikes have been constructed along the main rivers.
Table 2-2: Details of embankments and spurs at the provincial and national levels
Embankments (km)
Spurs (no.)
Total in Pakistan
Source: FFC annual report 2008
Flood protection bunds have generally been constructed either to protect headworks,
irrigation structures, or certain towns and villages. Controlled breaching of embankments
is also practiced to avoid unwanted breach. In the upper Indus Basin, the main rivers flow
in a south-west direction. The general slope is southwards, meaning that most of the
canals stem from the left banks of the rivers. Breaching is usually produced on the right
banks to avoid devastation, as most of the development is also on the left side where the
canal irrigation system is located. A double line of flood embankments have been
constructed along (almost) both banks of the Indus in Sindh province stretching from the
Guddu Barrage to a few kilometers before the river forms its delta. The embankments
have been further compartmentalized to contain inundation.
Floods in the upper reaches of the Indus and Jhelum rivers have been attenuated since the
construction of the Mangla and Tarbela dams in 1967 and 1974, respectively. Though the
storage capacities of these dams are decreasing due to sedimentation, they still play an
important role in flood management. The useful lives of these dams are expected to expire
in 2050 and 2060 for Mangla and Tarbela dams respectively (MoWP, 2002c; Izhar-ul-Haq
and Abbas, 2008; Hashmi et al., 2009). Their effectiveness in flood control is subject to
their storage capacities, adopted reservoir operation practices, and intensities of floods.
Although these dams are multipurpose, their prime function is to store water for
irrigation and power generation. The operation planning of these dams has not yet been
optimized to control floods downstream.
The Mirani Dam was constructed in 2006 on the Dasht River for the storage of hill torrent
water for irrigation purposes in Baluchistan. It enabled irrigation supplies on both sides of
the river and minimized flood damages in the floodplain (Majeed and Khan, 2008). About
12 sub-projects of protecting bunds and delay action dams were constructed in
Baluchistan under FPSP-II (Contijoch, 2008). The harnessing of hill torrents in D.G. Khan
has also been studied by the National Engineering Services Pakistan (NESPak) in 1984
and by the Japan International Cooperation Agency in 1992 (MoWP, 2002b). NESPak
accomplished another countrywide feasibility study on hill torrents in 1998. The study
area was divided into 14 hill torrent zones in the Federal Areas, NWFP, Punjab, Sindh, and
Baluchistan (Figure 2-1). Structural work has been completed in a few sub-zones of D.G.
Khan (e.g., Kaha and Mithawan).
Non-structural measures
All the major rivers in Pakistan are transboundary and flow through India. The shape of a
flood wave mainly depends upon water management practices in the watershed and
upstream operations. Being a low riparian country, flood management options are limited
and flood prediction is complicated in Pakistan. Therefore, main emphases have been put
on precise flood forecasting and an early warning system. Flood warning is mainly the
responsibility of the Flood Forecasting Division of Pakistan Meteorological Department
but the Water and Power Development Authority (WAPDA) also contributes to improve
the ability to forecast. The flood early warning system was initiated in 1975 when a realtime VHF telemetry system was introduced for hydrological data collection from 16 river
gauges and 24 rain gauges (Figure 2-3) (NESPak, 2008). A total of about 40 stations were
established at all rim stations and within the Mangla Dam catchment area. The number
was gradually reduced to about 20 due to maintenance problems. The Flood Early
Warning System (FEWS) was updated under FPSP-II in 1998 in cooperation with the
NESPak-Deltares Consortium. FEWS is a physically­based hydrodynamic model using
real-time data. The meteor-burst based communication system was integrated into the
FEWS through the WAPDA’s “Surface Water Hydrology Project” in 1998. About 22 high
frequency radio sets were installed to serve as a double support for automatic gauging
and the telemetry system (ADB, 2008). The high frequency radio system works as a
backup for telemetry and the meteor burst system.
Currently, flood-zoning considerations do not exist in Pakistan. Development of flood risk
mapping for the main rivers was initiated under FPSP-II. So far, hazard maps for 5-years
and 50-years return periods have been compiled. Calibration and risk assessment of these
maps is planned in the forthcoming NPFP. Interpretation and legislation regarding flood
zoning will be carried out afterwards.
The larger and more productive part of the flood-producing upper catchments of the
Sutlej, Ravi, and Chenab rivers lies across the border in Kashmir (Indian) (Figure 2-3).
Precise and timely measurement of precipitation in those areas is critical for effective
functioning of FEWS. A weather radar unit at Sialkot was installed with the ability to
detect the position of clouds and precipitation within a radius of 230 km. This radar
covers catchment areas of about 17 tributaries. A 10 cm S-band Doppler Weather
Surveillance Radar unit, installed in 1997 at Lahore, provides rainfall data about the
Sutlej, Beas, Ravi, and Chenab catchments from across the border (NESPak, 2008). Floods
in the Jhelum River occur mainly due to heavy rainfall with very short lead-time.
Therefore, a weather radar unit at Mangla was put up during FPSP II to provide
quantitative rainfall forecasts. More radar units have been planned to cover the hill
torrent generating catchments of D.G. Khan, NWFP, and Baluchistan.
Figure 2-3: Locations of telemetric gages, HF radios, weather radars, and river structures
The Mangla and Tarbela dams were constructed for irrigation and power generation
operations. Current reservoir operation practices do not play any substantial role in flood
management. The clear example is the recent flooding 2010, in which the Tarbela dam did
not play any significant role in reducing flooding downstream. Improved reservoir
operation of the Mangla dam to facilitate flood management was included in FPSP-II, but
now has been postponed due to a Mangla dam raising project. Pre-flood releases on the
basis of the flood forecasts can create required flood storage capacity. Improved planning
of reservoir operations for the Mangla and Tarbela dams is included in the next NFPP to
enhance their role in flood management.
The Pakistan Meteorological Department issues daily satellite cloud pictures from the
polar orbiting meteorological satellites on its website to inform the general public. In case
of cyclones, warnings are issued quickly. Cyclone detection radar is used for tropical
cyclone monitoring. Japan has donated radar equipment to the WMO regional center for
Bangladesh and Pakistan. This radar had contributed substantially to the detection,
monitoring, and forecasting of tropical cyclones in the country. Pakistan is a member of
the WMO and the ESCAP (Economic and Social Commission for Asia and the Pacific) Panel
on Tropical Cyclones, which aims to promote measures to improve tropical cyclone
warning systems in the Bay of Bengal and the Arabian Sea. A technical plan aimed at the
development and improvement of the cyclone warning system in the region has been
drawn up by the panel (WMO, 2008).
In addition, a number of control structures have been constructed in India, making the
operation of rainfall or runoff models more complicated. An agreement was signed in
1989 between the two countries to share river flow and rainfall data for flood forecasting
(Awan, 2003). The ‘zero flood warning’ manual was also accomplished to homogenize the
flood warning procedures and emergency action plans under FPSP II (Awan, 2003; FFC,
2007). Tackling the flood problem within flood managing bodies seems to become a
smoother and better-organized process. The setting up of standard operating procedures
may produce better interagency cooperation and coordination.
Legal framework
According to the Constitution of Pakistan, water is a provincial government responsibility,
but the federal government also performs a number of tasks and responsibilities in the
water sector, mostly relating to international and inter-provincial matters. The federal
government, through the WAPDA, the Indus River System Authority (IRSA), and the FFC
performs coordinated planning, development, and management of water and hydropower
resources. The legal framework for carrying out these tasks is provided by the WAPDA Act
(1958), the Environmental Protection Act (1997), the Indus River System Authority Act
(1992), and by the Constitution under various articles on inter-provincial coordination
and resolution of conflicts through the Council of Common Interests.
Recent policies dealing with crises are the Emergency Services Ordinance (2002) and
National Disaster Management Ordinance (2006), which provide the national strategy for
dealing with emergencies. A Draft National Water Policy by the Ministry of Water and
Power (MoWP) in 2002 was prepared to address most of the water-related issues in the
country, including flooding. This policy emphasizes all necessary structural and nonstructural measures for flood management and the need for stakeholder participation in
the flood management process, as well as enhanced flood awareness in the community. It
also recommends replacement of various water-related acts with a simple unified law that
enables clearer understanding and subsequent application of the law (Rehman and
Kamal, 2005). A number of strategies, visions, initiatives, and plans have also been
prepared, including the Ten Year Perspective Plan (by the Planning Commission in 2001)
and Vision 2025 (by the WAPDA in 2001) etc.
Pakistan has a very important agreement with neighboring India. The partition of the
subcontinent created a conflict over the water distribution rights of the Indus Basin. This
trans-boundary water issue between Pakistan and India was addressed with a temporary
‘Standstill agreement 1947’, the ‘Inter-Dominion Accord 1948’, and eventually the ‘Indus
Water Treaty’, which was signed with the help of the World Bank in 1960. Six main rivers,
the Indus, Jhelum, Chenab, Ravi, Beas, and Sutlej, along with their tributaries, are covered
in this agreement. According to this treaty, the exclusive rights of water use for the three
western rivers (Indus, Jhelum, and Chenab) were given to Pakistan and rights for three
eastern rivers (Ravi, Bias, and Sutlej) were awarded to India. Compensation to the eastern
rivers was managed with a number of link canals.
Institutional arrangements
Many federal and provincial institutes are involved (directly or indirectly) in flood
management activities. Based on the nature of services and support provided, these
institutes can be grouped under risk-managing and crisis-managing institutes. Riskmanaging institutes deal with prevention and relief, whereas crisis-managing institutes
are concerned with rescue, relief, and rehabilitation operations.
Hazard managing institutes
The Federal Flood Commission was established in 1977 and assigned the task of
preparing the NFPPs on a countrywide basis. Their specific jobs are to construct flood
protection and river training works, improve the weather radar data collection system,
and create awareness and adaptability among the local population. The FFC has played
the main role in the country’s flood management since 1977. Normally, flood protection
schemes are prepared by provincial governments (Provincial Irrigation and Drainage
Authorities) or concerned federal agencies. These schemes are then reviewed and
approved by the FFC, either on an emergency basis or in the context of a group of projects.
Flood protection plans in Pakistan are prepared on a countrywide basis by consultants
under the supervision of the FFC. Funding is provided by the FFC and execution of these
projects is carried out by provincial agencies. The FFC monitor and evaluate these works.
These projects can be executed as an individual independent project or as a subproject of
the NFPP.
The approach followed by the FFC encompasses both structural and non-structural
measures. Non-structural measures mainly pertain to the establishment of modern flood
forecasting and warning systems to provide timely and reliable flood information to the
flood mitigation agencies and to the public.
The Provincial Irrigation and Drainage Authorities (1997) are an upgraded form of the
Provincial Irrigation Departments with the extended scope of irrigation and drainage
management. The Provincial Irrigation and Drainage Authorities play an important role in
flood mitigation by performing design, construction, and complete maintenance of river
training and flood protection works. These also provide the flow measurement of rivers,
canals, and drains for flood forecasting. In addition, their role in crisis management is to
prepare flood emergency plans before, during, and after the floods.
The Water and Power Development Authority is involved in the flood forecasting
process by providing river and rain data from its telemetric gauge sites within the upper
catchments of Indus and Jhelum rivers. The safety of the Mangla and Tarbela dams are the
top priority for this data collection. It is also involved in providing inflow and outflow data
from the Mangla and Tarbela dams and the Chashma barrage.
The Flood Forecasting Division of the Pakistan Meteorological Department collects
hydro-meteorological data from various national and international sources and then
processes data to produce flood forecasts and warnings. Flood warning dissemination is
solely the responsibility of the chief meteorologist to avoid rumors and misinformation
about floods.
Crisis managing institutes
Crisis management is mainly performed through a set of administrative entities.
Therefore, it will be convenient for international readers if administrative divisions in
Pakistan are described before discussing the existing institutional setup. The country is
divided into 5 provinces each having their own political government. These Provinces are
further divided into Divisions that, in turn, consist of Districts. Both Divisions and Districts
are only administrative levels headed by Commissioners and Deputy Coordination
Officers without political representation. Each district is further divided into Tehsils and
Tehsils into Unions that are represented by elected Councilors.
The Provincial Relief Departments are responsible for flood preparedness as well as
rescue and relief plans. The department arranges surveys to ensure that all flood
protection bunds are satisfactorily maintained before the flood season. It sets up flood
warning centers and flood centers at district and union levels. In fact, the Relief
Department functions through control and coordination of the personnel and resources of
other government departments generally organized in form of committees.
The Emergency Relief Cell works at the federal level and mainly deals with the planning
and assessment of relief requirements for major disasters. The scope of their activities
covers stock piling of necessities needed during an emergency, establishing emergency
funds, and assisting international donors in their relief efforts. The provincial
governments and local administrations provide relief for disasters. The National Disaster
Plan from 1974 covers procedures, organizational set-up, and standard procedures for the
monitoring of disaster operations.
The Army provides necessary help to civil authorities to carry out rescue and relief
operations during and after floods. The Army also takes part in pre-flood season surveys
and inspections of the flood protection works. It is the responsibility of the provincial
government to provide all support equipment (boats, life jackets, vehicles, tents, etc.) to
the Army for these operations. During the flood season, the Army sets up flood emergency
cells at each corps headquarters. In the case of major floods, the Army is responsible for
actuating controlled breaching of pre- defined flood dikes to divert the peak away from
the cities. Although, there exists no standard procedure, the breaching is decided on the
basis of existing and forecasted flood stages with the mutual consultation of local officials
of civil administration, irrigation department, and army. The Army has been playing a
vital role in flood relief activities in 2010 flood since the start of this disaster. Their relief
activities demands intense cooperation with organizations that provide flooding
information. There are also a number of departments, which are assigned special tasks
during floods.
Figure 2-4: Flood losses details at national level against severe flooding years, Sources:
FFC annual report 2008 and National Disaster Management Authority
Analysis and discussion
The overall data of lives lost and villages flooded (Figure 2-4) shows a decreasing trend
from 1950 to 2009, which may be due to improvements in flood management. According
to the Centre for Research on the Epidemiology of Disasters - International Disaster
Database EM-DAT (1980–2000), the ratio between the number of deaths and population
exposed to floods in Pakistan is lower than Afghanistan, Bangladesh, India, and China
(Pelling et al., 2004). Whereas flood losses at the worldwide level demonstrate an
increasing trend (Pielke, 2006), flood losses in Pakistan showed a decreasing trend due to
improved defense until the recent flood. The sense of safety induced by the decrease in
floods resulted in increased vulnerability of society. As a result, life losses and financial
losses were exceptional during the 2010 flood, given the flood levels, which were the
same as in 1978 at Kalabagh gauging station (Figure 2-2 and Figure 2-4). Main factor was
poor maintenance of dikes.
Flood loss data at district level show that historic fluvial floods of the major rivers seldom
claim lives, whereas regular annual losses are mainly agricultural. Total areas flooded and
flooded cropped areas can be used as good indictors to assess the impacts of flood
management at district level. Therefore, flooded areas and crop areas flooded at district
level have been charted for major rivers upstream from the river confluence (Figure 2-5)
to evaluate trends in flood losses. Some reductions in the flooded areas have been noticed,
overall. Historical trends show that the country observes alternate flood rich and flood
poor periods. It is also worth noting that there has been no major flood since 1995 and
that the flood in 2010 occurred after a prolonged dry spell.
Though both structural and non-structural measures have been implemented to reduce
flood losses, available statistics show that flood management in Pakistan basically
revolves around structural measures with a primary focus on flood prevention (MoWP,
2002c). Crisis management strategies are mainly comprised of rescue and relief actions.
However, no solid strategy has been developed to enhance the flood fighting abilities of
individual communities. Flood mapping has been initiated but still no final and authentic
product has been produced to integrate flood mapping into existing flood management.
New initiatives for structural and non-structural measures are taken continuously but
lack of continuity and maintenance mostly results in failure. Poor maintenance of
telemetric system, dikes, and FEWS are among the examples. Dike failures and
malfunctioning of FEWS for flood warning due to poor maintenance and negligence have
been observed during 2010 flooding (Tariq and van de Giesen, 2010). The lack of social
support for technical designs also plays an important role.
Figure 2-5: Area flooded and crop area flooded at district level for major rivers, Sources: District governments and Punjab Irrigation department
Funds are controlled and provided by the federal government through FFC and there is no
consideration in terms of ‘who pays and who benefits’. On the other hand, the project
approval guidelines set by FFC (FFC, 2009) carry strategic biases that are aimed at
protecting locations and infrastructure of greater economic, political, and strategic
significance, at the cost of areas and communities with lesser influence and importance.
For a project to qualify the acceptance criteria, it must have an EIRR above a threshold,
usually set by donor agencies. Self-reliance and risk-based approaches are not yet part of
project acceptance criteria.
The social and economic infrastructure of Pakistan depends on the waters of Indus Basin.
Alarming records of historical flood losses (Figure 2-4) show the seriousness of the flood
problem. Measures have been taken for flood management, but there is no serious effort
to increase the system’s ability to cope with the fluctuations in annual and seasonal flows
in the Indus River System. Pakistan’s current water storage capacity is around 12% of
annual availability. No major dam has been constructed since the completion of Tarbela
Dam in 1974. Construction of new dams and reservoirs has been hindered by interprovincial disputes. The country was suffering severe draught and water shortage shortly
before it was hit by the devastating flood in 2010.
2.10 Discussion
Flood management in Pakistan is a task that requires both the resources and
comprehensive understanding of the flood problem. The nature of floods varies
drastically throughout the country due to contrasting physiographic, climatic, hydrologic,
demographic, and socio-economic factors. The present approach for flood management
incorporates both structural and non-structural measures, yet their inter-linkage and
combined efficiency still need to be optimized. The efficiency of any proposed measure
should be evaluated for its integration into existing measures to achieve efficient and
economically viable solutions.
Change in flow regime due to low flows in eastern rivers after the Indus Water Treaty and
enhanced flood protection measures have attracted economic activities and settlements in
floodplains. Flood management arrangements are concentrated around the Chenab and
Jhelum rivers because of the frequent and devastating nature of flooding. Those
floodplains that have not faced flooding over a considerable time are under extremely
high risk. Vulnerability on such locations has increased due to a false sense of safety. The
2010 flood in the upper Indus was due to exceptional intensive rainfall in the catchments
of the Kabul and Swat rivers, which was not covered by Doppler Weather Surveillance
Radar units. The Doppler Weather Surveillance Radar network should be extended to
cover northwestern areas of the Indus Basin to enhance the capability and reliability of
FEWS and the same system should be established for the hill torrent areas of the Kharan
Basin after carrying out feasibility analysis.
Currently, there exists no well-defined criterion to initiate new measures. Political
processes and influence shape flood management planning. The situation worsens, as
funding is not a responsibility of floodplain inhabitants. A race to secure more measures is
unavoidable. In addition, the protection of high value areas at the cost of low priority
areas promotes unlawful breechings of dikes, which was also observed during the flood in
2010. To overcome the problem, the risk-based approach must be incorporated to handle
flood problems within available resources. Resources required for flood management
must be generated from water users and floodplain inhabitants and dependency on
donors must be avoided. Comprehensive standard operating procedures must be
formulated based on risk and self-reliance.
Expansion of structural and non-structural measures is extremely important to enhance
the efficiency of the flood management system. Flood zoning and flood mapping projects
must be completed on priority basis. Necessary legal and institutional support must be
provided to flood mapping and flood zoning. New dams are necessary for improvement in
water management in general and for effective flood management in particular.
Unfortunately, maintenance and functioning of flood measures have been neglected. High
priorities must be assigned for the proper functioning of measures. FEWS is a state of the
art model. Its proper functioning and full utilization must be assured. Comprehensive
flood management plans must be prepared and executed without waiting for another
devastating flood.
Concluding, a risk-based pro-active approach is required to achieve sustainable flood
3 Risk-based assessment of flood
Neither the problem, nor its solutions, is new, yet with ever-growing environmental
awareness, standards of life, and expertise in technology, the demand for more effective
and efficient flood management is quite natural. In addition, flood management is a slow
dynamic process and should be updated on a regular basis, roughly 30-50 years (Plate,
2002). New approaches, ideas, and measures have been introduced accordingly. With
growing numbers of new terminology and concepts, three major concerns are raised:
Are there any new concepts with new terms, or just old concepts with new
terms? Do these new concepts replace old practices, or do these need to be
implemented alongside?
What are the characteristics of good approaches for flood management?
How can we come up with efficient and effective flood management practices?
These questions are addressed in this chapter. Basic components of flood management
are also defined for the understanding of important flood management concepts,
practices, and ideas. General guidelines for an ideal flood management strategy are
described. Basics of risk-based assessment are defined. It is elaborated how risk-based
assessment ensures a strategy that will meet the standards described in guidelines (See
Flood impacts
The nature of flood impacts determines the shape of flood management strategies. Floods
can have both adverse and beneficial impacts (Ref. Table 3-1). Flood impacts can be
summarized as all the effects that floods have on their environment including the
drowning, wetting, erosion, deposition, disturbances, rehabilitation, insurance,
management, etc. There is a wide variety of positive and negative flood impacts (De
Bruijn, 2005). In eastern Spain, for instance, large floods are very important for the
recharge of groundwater aquifers used for agriculture and for tourism, and for the
maintenance of coastal wetlands (EEA, 2003). Different flood impacts can be categorized
based on the following criteria:
Whether the impact is positive or negative
The connection/ relation with the flood
Direct: Impacts due to physical contact with flood itself (Dutta et al., 2001;
Merz et al., 2004; Nascimento et al., 2006)
Indirect: caused by the direct impacts through interruption and disruption
of economic and social activities and may occur in space or time outside the
flood event (Merz et al., 2004; Veerbeek, 2007)
o Primary: Indirect impact within floodplain
o Secondary: Indirect impact outside floodplain
Induced: Impacts related to efforts in context with flood management
The conventional expression of impacts in monetary values
Tangible: The impacts can be expressed in monetary terms using
conventional methods (Smith and Ward, 1998; Dutta et al., 2001)
Intangible: The impacts cannot be expressed in monetary terms using
conventional methods. They have an effect equivalent to 50% to 100% of
the direct financial losses (ANFAS, 2003).
Damage categorization determines which damage types are most relevant in a specific
situation. For the optimization of flood management, it is of utmost importance to
consider the types of damages and effective measures to take into account.
In developing countries, mostly the poor population lives in floodplains (Dilley et al.,
2005). On the other hand, it is also true that the people living in floodplains might become
poor as they are frequently flooded (APFM, 2009a). The benefits relating to agricultural,
ecological, environmental, groundwater recharge, and business activities about flood
management can be considered as positive impacts of floods. Based on these criteria,
examples for flood impacts are given in Table 3-1.
Components of flood management practices
Flood management practices vary by time and location. Present day practices differ across
countries and have evolved over time. These practices depend on the severity of the flood
problem, available resources, economic growth, and the social apprehension of flooding
and water resources. Recently researched policy documents new ideas, approaches, and
visions are often expressed by ‘concepts that seem to have turned into buzzwords’ (De
Bruijn, 2005). However, there remains a need to understand this terminology and the
basic structure of flood management. Following are the basic components of flood
management practices and conceptual composition are shown in Figure 3-1 and discussed
in more details in the following sub-sections.
Assessment/ Design criteria
Plans/ Projects/ Strategies
Table 3-1: Classification of flood impacts (river floods)
discoveries. Flushing of
salt from land surface
Victims, ecosystems,
monuments, culture
Increased business & Production
production for relief & income loss, theft and
rehabilitation industry robbery
Replenishing lakes and Social
fishing emotional damage
production. Groundwater
Increase in production
competitors outside
Occlusion of seawater
intrusion in estuaries and
coastal areas. Aquifer
recharging of outside
High nutrient water to Capital loss (houses,
crops reducing water crops,
and fertilizer costs
buildings), deposition of
pollution and debris or
Production losses for
supplier from outside the
unemployment, inflation
diseases to outside
area, migration
for Costs for relief aid, flood Raised
& Evacuation
flood protection & protection measures and regional cooperation in Land-use
insurance business
all management costs
relief and rehabilitation
Figure 3-1: Schematic concept of Measures, Approaches, Design Criteria, and Projects
Flood management measures are the measures taken to reduce the flood problems. This
way, flood management can be considered as coordination and management of these
measures (ADPC, 2005). New measures become available with the advancement in
science and technology. The suitability of a measure to any specific flooding problem is
subject to flooding behavior, available resources, and technical limitations at hand. The
selection of flood management measures is carried out in two steps. In the first step,
measures are shortlisted based on adopted flood management approaches (Ref. 3.2.2),
available resources, and technical suitability. In the second step, positive and negative
effects are evaluated according to assessment method or ‘design criteria’ (Ref. 3.2.3), and
the technical design is carried out.
There exists no single solution that can be applied everywhere. Consequently, it is not
logical to argue for or against any measure but a measure can be most suitable depending
upon the local situation (APFM, 2009a). That is, the suitability of a measure depends upon
both the socioeconomic conditions of the country and the behavior of the floods (Green et
al., 2000). Following are some common examples of flood management measures with a
brief description of their functionality:
Reservoirs reduce flooding by holding or attenuating flood wave peaks.
Rain harvesting stores part of the runoff for agriculture or domestic use and
reduce flood volumes.
Dikes, levees, floodwalls, and other barriers are erected between a river
(source of flooding) and the settlements thereby to avoid floods.
Channel improvements increase the flow capacity of a stream/ channel by
making it wider, deeper, smoother, or straighter.
A diversion (or by-pass) is a channel that is used to divert part of the peak
flow and may return it back at downstream.
Insurance and relief are traditionally considered as tools for recovery after
flood occurrence but may also play an indirect role in flood risk control.
Flood warning, rescue, and pre-emptive evacuation help by timely moving
lives and goods out of the floodplain.
Public awareness is highly flexible and extremely effective measure.
Providing people with a basic understanding about floods, flood
management, and emergency responses can reduce their vulnerability
Flood zoning, encroachment control, and implementing building codes in
floodplains reduce vulnerability and hence reduce flood losses.
Details showing the role of some important flood management measures and their
classification have been described in Table 3-3.
Flood management approaches can be defined as the way to deal with flood problems. It
can also be considered as a long-term planning exercise (Halcrow, 2004). Similar to
measures, the appropriateness of a flood management approach depends on the context
of the floodplain and country (Green et al., 2000). No flood risk management approach is
superior in all aspects and in all conditions (Middelkoop et al., 2004). The development of
a suitable approach needs to consider the socioeconomic and environmental conditions as
well as the severity of flood problems at national level. Unfortunately, when developing an
approach, a simple ‘copy-paste’ function that is being practiced in developing countries,
will not work. Green et al. (2000) describe such practice as “Simply proposing to adopt the
approach that is appropriate in one area to another area is parochial at best and neocolonialist at worst”. Flood management approaches of a country represent the priorities
of social, environmental, and economic assets.
Flood control and flood mitigation approaches tend to confine flood spread
through structural measures.
Adaptation is another concept that describes the adjustment or modification
of human activities to minimize flood losses.
Resilience-based flood management is an approach that emphasizes to
structure floodplain activities such that the system can recover after flood
Integrated flood management emphasizes connecting flood management
and other river functions and floodplain activities appropriately.
Multidisciplinary expertises are involved.
Sustainable flood management ensures the selection of such measures that
do not cause grave complications in the future. This approach demands a
fair valuation of social, economic, and environmental assets.
No adverse impact approach claims the formation of such plans that do not
shift or increase problems in adjacent areas.
Floodplain restoration is based on the understanding that a natural
floodplain has better capabilities to handle floods.
Although, some approaches are widely appreciated and practiced, few failed to attain
worldwide attention. Setting fascinating targets and using appealing terminologies to
obtain acceptance without considering ground realities is one of the reasons that made
few approaches unpractical. Selecting a suitable approach is subjected to thorough
assessment for its suitability at national level.
Assessment/ approval criteria
‘Assessment’ can be defined as the set of minimum required standards or specifications
that must be met in order for a specific measure to be selected. Alternatively, it can be
defined as the procedural evaluation of positive and negative impacts of any measure or
plan. Measures that are short-listed on the basis of the adopted approach are further
analyzed for their suitability, along with their design specifications. Measures provide
safety but cost money. Some measures have environmental and social consequences in
addition to economic expenses. Thus flood risk assessment is a tradeoff between risk and
financial investments (Middelkoop et al., 2004).
Assessment criteria should be clear, transparent, and objective oriented. These may
include social appraisal, economic evaluation, and environmental assessment. In addition
to assessment criteria, every individual measure is designed on its technical grounds.
Non-structural measures need detailed assessment of their impacts under institutional
planning. Measures that involve engineering structures are designed under hydrologic,
hydraulic, and structural prospects.
The development of assessment methods has been a continuous process. At present, a
number of assessment methods are used in practice. Following is a brief introduction of
these assessment methods:
Element-design standards
Probability-based designs
Risk-based assessment
The purpose of all the standards, developed so far, is weighing benefits and costs of
measures (Hoes and Schuurmans, 2006). The improvements that are made in assessment
approaches are due to the better understanding of flood management.
Element-design standards
Element-design standards can be considered as the very first type of standards brought
into practices. The purpose of incorporating such standards was to establish a common
understanding, and to set the minimum quality that must be ensured. These standards are
derived from experiments or (successful) past experiences. According to this approach,
the design of flood protection structures must follow in-practice minimum standards.
Heights of flood protection structures are designed against historical flood experienced
(Andjelkovic, 2001; ASFPM, 2004) or, alternatively, design heights are set as multiples of a
rounded figure, say 5ft.
These standards require fewer efforts for analysis. However, they are not optimized in
terms of the costs to benefits ratio. These standards are still used in practice in those
cases where high accuracy is not strictly required (Nathwani et al., 1997). One form of
such standards remained in practice by the Connecticut Resources Commission in the
USA, that used 5-7 times the mean annual flood as a standard as late as the 1960’s
(ASFPM, 2004). Until 1953, dikes in the Netherlands were constructed to withhold the
highest known water level without being overtopped (Roos and Jonkman, 2006).
Probability-based standards
In the field of hazard and disaster management, safety is given prime importance. With
the advancement of the knowledge of disaster management, probabilistic safety standards
are developed and implemented worldwide. In the probability-based approach, the
degree of flood-control is expressed by the return period of flood (N-years) (Duivendijk,
1999). The probabilistic approach tends to assume that events in the future can be
predicted based on the extrapolation of past observations (Pistrika and Tsakiris, 2007). At
present, these safety standards are the most widely practiced assessment methods to
design flood management plans.
By the early 1960’s, a uniform standard (100-years probability) for the design of flood
protection structures was recognized in the USA (ASFPM, 2004). Currently, this standard
is still used to identify, map, and manage flood hazards (Carter, 2005; Kron, 2007). In
many other countries, such as United Kingdom, Germany, Italy, Spain, France, Canada and
New Zealand, a 100-years flood plays an essential role in flood mitigation strategies (Apel
et al., 2009). Similar standards are followed practically worldwide. For example, the
Netherlands have standards as high as 10,000-years against sea floods for the province of
Holland (J.K, 2001), whereas Bangladesh follows a design standards of 20-years
(Duivendijk, 1999). Pakistan adopts 50-years for designing flood protection structures
and 100-years for vital river training structures and bridges (Halcrow et al., 2001).
Establishing a design probability and defining an acceptable risk always remained a
matter for debate. Unfortunately, instead of defining the safety standards on the basis of
the acceptable risk, in practice, it is most often the reverse. For example, Hunter and
Fewtrell (2001) relate the acceptable risk and the probability standard as “a risk becomes
acceptable when it falls below an arbitrary defined probability”. However, such standards
target only floods of the designed probability. These standards lack the ability to deal with
severe floods, and the response would typically be an ad-hoc reaction in case a severe
flood occurs (APFM, 2009a). Another shortcoming of this method is that only the hazard
probability is considered irrespective of consequences, while the advantages of a measure
depends on the damage prevented (Hoes and Schuurmans, 2006).
The probability-based approach delivers more information than Element-design
standards that may develop a false sense of complete security. The probability-based
standards come up with the indication that there is always residual probability of failure
of a measure. On the other hand, just like the Element-design standards, this approach
may lead to inefficient use of precious resources or insufficient protection due to ignoring
the exposure (Chetty and Smithers, 2005) and susceptibility of the society. Optimization
of the flood management measures is not possible using these standards.
Probability standards and acceptable risk are related to location characteristics and
cannot be the same everywhere (Hunter and Fewtrell, 2001). Most developing countries
follow the probability standards adopted by developed countries due to the lack of
research, knowledge, and/ or the fact that most of such projects are designed and
executed with foreign help. These assessment criteria seldom fit to their socio-economic
Risk-based assessment
Until recently, flood control and protection have been engineering-centered and based on
probability. Little or no attention was given to the social, cultural and environmental
effects of any specific strategy (APFM, 2006a). In the field of flood management, riskbased assessment methods were introduced during the 1990’s. The main characteristic
that distinguishes this method from the previous two methods is its clear focus on flood
impacts, instead of floods. The risk-based assessment tries to reduce all possible floodrelated risks rather to consider the impacts of a specific discharge (Duivendijk, 1999). In
the risk-based design, the design return period is a decision variable and not a preselected design parameter value, as with the probability-based design method (Tung,
2002). The flood-risk at a location can be defined as the potential damages due to all
possible floods. These damages can be economic, social, and/ or environmental. Here,
tangible losses are comparatively easy to handle while advanced methods are needed to
evaluate intangible losses in monetary terms. A cost-benefit analysis forms a core
component of the risk-based assessment. In case of floods, there are many benefits
involved. This method involves extensive data analysis and computer processing, but has
the ability to optimize the flood management.
Design of a flood measure contains a number of uncertainties. Assessment methods are
means by which to judge whether a measure should be selected to achieve a specific goal.
All of the above-mentioned assessment methods have uncertainties at different levels, e.g.,
in data processing and target orientations. Even though the uncertainties remain there,
decisions are made based on expected outcomes (Nathwani et al., 1997). In this thesis, the
risk-based assessment method has been applied with an improved assessment model
(Figure 3-2) for the optimization of flood management measures. This improved model
may significantly enhance the overall efficiency of flood management.
Plans/ projects
Plans or projects are the outcomes of a flood management strategy. A plan consists of one
or more measures. A project is a practical form of (part of) a plan. Depending on the
severity of the flood problem, plans are prepared. The availability of required resources
transforms plans into projects. It is possible that a combination of measures for flood risk
management can be implemented throughout the lifetime of a strategy (Halcrow, 2004).
The worldwide practice of flood management is mostly reactive rather than proactive.
After each severe episode of flood, governments incur considerable expenditures directed
at the plans for flood management. Practical examples are the establishment of the
Afsluitdijk as a result of the Zuiderzee flood in 1916, the Delta Commission after the 1953
flooding in the Netherlands, and the Federal Flood Commission of Pakistan (Ref. 2.8.1) in
1977. Both Commissions were established for long-term comprehensive flood
management plans for their respective countries. ‘Room for the River’ and ‘Delta Plan’ in
the Netherlands and ‘National Flood Protection Plan’ in Pakistan are the examples of
large-scale long-term plans.
Guidelines for flood management approach
As mentioned at the start of this chapter, new strategies and measures are both needed
and become available over time. New measures and approaches must be considered,
assessed, and compared with measures adopted in the past. Before choosing flood
management options, the policy makers must envisage the consequences thoroughly.
Different countries have a different nature of flood problem and different capabilities,
therefore the solution should be different (Ref. 3.2.2). Just creating flood storages,
protecting dikes, or abutments in floodplains are not economically feasible measures
everywhere. The developing countries usually lack research, and continue to replicate the
flood management practices in developed countries. In addition, the act of imitating the
flood management practices of rich countries while having insufficient resources creates
problems. Every country must develop its own indigenous strategy according to the flood
behavior and available resources. In following sections, important guiding principles are
described to shape the national flood management strategy of a country.
Enhancement in flood management
The most important guideline is that a strategy can only be adopted if it enhances the
flood management. The word ‘enhancement’ means reducing the flood risk without
compromising social, economic, and environmental aspects of overall progress. Trends in
national flood losses need not necessarily provide any guidance to the success or failure of
the adopted strategy (Green et al., 2000). Reduction in risk due to better flood
management attracts more economic activities to the floodplain. In such situations,
though a rise in both the losses and the costs of flood management can be observed, yet
the economic gains might be higher even after compensating these costs. In other words,
the damages with the project should be less than the damages without it, or simply, the
net benefits must be positive (Medina, 2006). Therefore, the enhancement in flood
management must be assessed by the relative increase or decrease in the overall risk.
Nathwani et al. (1997) emphasize incorporating the overall impacts: “it is foolish to seek
maximum benefit without considering the risks involved, but it is just as foolish to pursue
minimum risk without regards for the cost”.
The desire of full protection prevailing a few decades ago, has now shifted to a concept of
risk-based sustainable floodplain management, as it has been realized that risk cannot be
entirely eliminated in many cases (Pilon et al., 2003). The majority of flood management
options redistribute the risk spatially and temporally. This redistribution can be
considered an enhancement when the overall risk is reduced, which is described by the
‘Kaldor-Hicks Compensation Principle’ stating, “A redistribution of risk is efficient if it
enables the gainers to compensate the losers, whether or not they actually do so” (Ref.
3.6.1). In sum, a risk redistribution that results in risk reduction is an enhancement in
flood management.
Identification of problem
Floods should not be seen as the problem; it is the impacts of the flood that should be
considered as the actual problem. Floods can be beneficial in many cases where
floodplains are less occupied. Actual estimation of total losses is operose as flood damages
are both tangible (direct and indirect) and intangible (Andjelkovic, 2001). The initial
response against flooding was to prevent floods. Later on, it was realized that there is a
need to minimize the flood losses and increase the flood benefits. Therefore, a systematic
approach is required to identify the nature of the problem and those measures that can
reduce the severity of the problem more effectively.
Holistic considerations
Whether designing an individual measure or a strategy, both should consider all possible
impacts of all possible floods. Although a probability-based design possesses more holistic
approach as compared to element-design assessment is done against a specific return
period only. However, flood management can be more effective if it considers and
addresses all probable floods (Green et al., 2000; APFM, 2009a). Therefore, the whole
range of possible floods must be considered and addressed accordingly. A design
according to a specified probability standard, without carrying out detailed assessment of
all possible floods, is not technically sound and acceptable.
Integrated approach
Flood management is not an isolated process that is disconnected from the functions of
the river and society. An integrated approach is essential in order to find and evaluate
alternative strategies (De Bruijn, 2005). The International Commission on Large Dams
(ICOLD), defines four distinct successive periods in the approach development
(Duivendijk, 2006). From the time when people were not capable of doing much against
floods, to implementing structural measures, to incorporating non-structural measures,
and finally, adopting an integrated approach. Integration of flood management can be
defined by the following aspects
Integration with river processes
Integration with societal functions
Profit-loss sharing
Integration with river processes
To establish an implementable floodplain management policy to address the flooding
problem, in all cases, solutions must consider the whole functionality of the river. Rivers
perform multiple functions to sustain social, economic, environmental, and ecological
prospects of floodplain activities. Basin-wide consistency of flood management is
recommended for such integration (Duivendijk, 1999).
Integration with societal functions
Flood management will only be successful if it works with, rather than against, community
goals and priorities (Montz and Gruntfest, 2002). It is extremely important to consider the
behavior of society while designing the flood management strategy. After all, not the river
but the functioning of the (local) society is at stake (De Bruijn, 2005). Therefore, it directly
follows to consider societal functions and behavior, as well as river processes.
Profit-loss sharing
The majority of flood management measures reduce flood losses by redistributing these
risks. While redistributing these risks, the policy makers should not ignore the
distribution of the benefits. Issues of ‘who benefits and who pays’ often hamper source
control efforts (Andjelkovic, 2001). Strengthening the society to cope with floods, profitloss sharing relies on fairness and social justice in society. The decision-makers must
envisage whether the redistribution of risks in time and space is viable (APFM, 2009b).
Incorporation of diverse measures
Optimum flood management may incorporate many structural and non-structural flood
management measures with certain priorities and combinations. Introducing nonstructural measures along with structural measures is vital to reduce damage (Duivendijk,
1999). Not all measures can be applied to all locations, as their efficiency is subject to
flood characteristics, societal responses, interaction with other measures and their
technical designs. Every measure possesses specific characteristics and attenuates risk in
specific aspects. Understanding these characteristics of any measure makes it easier to
choose the appropriate flood management option fulfilling the local constraints. These
constraints could be economic, social, and environmental in nature.
As mentioned, the impact of a measure is not independent of flood conditions and other
in-situ measures. In addition, the technical design and planning of a measure has a large
influence on the efficiency of that and any other measures. Given the potential for synergy
in effectiveness, where possible, structural measures should be backed up with nonstructural measures. Non-structural measures need extreme care in planning; otherwise,
their efficiency may reduce severely and may even increase the risk instead. For example,
the new-for-old insurance policies have increased flood losses (Messner et al., 2007).
Utmost care is required while incorporating non-structural measures. The ultimate
impacts of non-structural measures depend upon a number of factors (Yoe, 1994). The
careful coupling of non-structural measures to structural measures may prove beneficial
and limit flood damage (Duivendijk, 1999).
Balanced approach
Flood management requires an adequate balance between the river processes and the
societal activities in the floodplain. Furthermore, flood management must establish a
social, economic and environmental balance using quantitative analytical tools
(Duivendijk, 1999; Young et al., 2000). Flood management should not be a practice merely
to suppress the floods, or the human activities. Traditionally, all risk was attributed to the
floods; this is the reason why the measures adopted for flood management in the past
(and are in use in most of developing countries and to some extent in many developed
countries) extremely suppress river functioning. Growing environmental awareness
changed the approach and points at the responsibility of human activities. Eq. 3-2 shows
the involvement of both sides.
Uniformity and fairness
In most cases, flood management is a government task. Establishment of a common
framework is necessary to guide decision-makers to prioritize flood management
uniformly over the entire country (Nathwani et al., 1997). An unbiased, uniform, fair, and
clear approach is required to implement flood management nationwide. Duivendijk
(1999) recommends the implementation of a uniform policy at basin level suiting the
current economic, social and environmental values. However, economic, social, and
environmental values may not be equal at basin level. Therefore, uniformity of approach
at national level and planning at basin level can be considered as appropriate.
Another common source of inequity is assigning temporal priorities to environmental
assets or to life-saving, while designing individual projects. Priorities to environmental
assets and life-savings must clearly defined be forehand at national level considering
social and economic constraints, and should be the same for all areas and all projects.
Long-term considerations
Policymakers must envisage the short-term and long-term impacts of a selected measure.
Every measure that is taken is an intervention, and may thus disturb the dynamic
equilibrium of the system by introducing some adverse impacts. Therefore, these impacts
must be envisaged by performing prior analysis. Experience has shown that structural
measures can cause severe impacts on the floodplain ecology. This results in increasing
flood losses, environmental degradation, and an inability to incorporate climatic changes
(Lyle, 2001). For sustainable flood management, it is important to choose an option that
does not compromise the future. The flood management should be flexible, adaptive, and
future-oriented both in short and long-term. We should have a preference for options that
involve resilient natural systems with enhanced but adjustable coping capacity of
individuals and communities (Green et al., 2000).
Risk-based assessment
Risk-based assessment relates to the evaluation of flood management measures based on
their potential impacts. These potential impacts are comprised of probable social,
environmental, and economic consequences, and are typically referred to as risk.
Consequently, flood management can be considered effective only when it reduces net
negative impacts of floods. On the other hand, risk-based analyses are the only systematic
procedure to achieve flood management optimization, due to their orientation towards
risk instead of flood. Recent research therefore suggests to follow this risk-based
approach in flood management (Hooijer et al., 2002; Hoes, 2006; Moel et al., 2009).
The concept of risk-based design has been around for some time, yet there still exists
reluctance and resistance in accepting practical problems (Tung, 2002). A risk-based flood
management approach provides the logical grounds for selection and design of flood
measures. It incorporates fairness, uniformity, and firm logical bases for flood
management practices. Based on risk, the proposed assessment method provides a
systematic approach that allows decision-makers to balance losses and benefits properly.
Before going into details on flood management, it is important first to understand the
mechanism how floods affect human developments. In order to accomplish a
comprehensive approach to risk-based flood management, it is necessary to make a clear
understanding of the risk concept. There is a certain need to elaborate the role of river
processes and societal activities in inducing risk. To understand the risk-based
assessment better, a detailed explanation is given in coming sections. In addition, basic
terms are also defined to allow for clarity and avoid ambiguity. Although there is general
agreement on most of these definitions, some scientists define a number of terms slightly
The term risk has been defined and understood somewhat in similar meanings
worldwide. Several aspects have been emphasized while being defined by different
scientists. In defining risk, some over-simplification or redundancy of terms is often found
in literature. The most commonly used and agreed upon definition of risk can be ‘the
expected losses due to a hazard’ (UNDHA, 1992; Smith, 1996; Crichton, 1999; Granger et
al., 1999; Kron, 2002; Sayers et al., 2003). Losses may comprise of lives, health, social
disruption, environmental impacts, economic losses etc. Others define risk as ‘the
exposure of something of human value to a hazard’ (Smith, 1996).
In pseudo-mathematical terms, ‘risk is the product of probability and loss’ (Helm, 1996;
Smith, 1996; Sayers et al., 2003) (Risk = Probability × Consequences). Kron (2002) defines
risk as the product of a hazard and its consequences. Some define it as ‘the product of
hazard and vulnerability’ (UNDHA, 1992) (Risk = Hazard × Vulnerability). While others
include the ‘hazard impacts, elements (value, exposure) and their vulnerability’ (Blong,
1996; Crichton, 1999), Cruz-Reyna (1996) include ‘preparedness’ as a dividend, Granger
et al. (1999) included ‘elements at risk’ in definition of risk.
It is obvious that risk has been understood in a similar way, yet major differences can be
seen when expressing risk mathematically. The main reason is that the terms
vulnerability and hazard are defined differently. To come up with consolidated concepts
and an agreed upon understanding of terminology (before explaining the application of
risk-based assessment in the next chapters) basic terms and the corresponding
expressions have been established here.
There is no doubt that risk is associated with hazard, probability, consequences,
preparedness, impact of hazard, value of elements at risk and vulnerability (where these
parameters have been defined also differently). Concluding, as a proper definition based
on generally agreed concepts, risk can be broadly conceived as ‘an estimate of the
potential consequences associated to a hazard’.
= Probability (year-1) × Consequences ($)
Eq. 3-1
= Hazard (m year-1) × Vulnerability ($m-1)
Eq. 3-2
Risk is a function of hazard, vulnerability and their mutual interaction.
The above-mentioned expression shows that hazard alone is not responsible for inducing
risk. This is the reason why measures that reduce the vulnerability of a society must also
be considered. There is a certain need to elaborate the role of both sides in inducing risk.
This understanding will help in optimizing the measures and land-use planning in the
floodplain by minimizing the risk for sustainable flood management. The exposure and
susceptibility of the vulnerable on one hand, and the intensity and probability of the
hazard on other hand, are the main parameters to determine risk in a quantitative way.
Everyday hazards and vulnerability form patterns of accumulating risk that can culminate
in a disaster due to an extreme natural event (Pelling et al., 2004). Therefore, we can
analyze risk from two independent factors: hazard, and vulnerability.
The ‘hazard’ in flood management is defined as the occurrence of a high water level event
with a defined exceedance probability (Kron, 2007).
In the context of risk management, the hazard is that which triggers risk once the defense
is exceeded. The hazard can be characterized by its probability and intensity. In an area
where the probability of a flood is practically zero there is no risk involved, and for an
area where a flood occurs, the losses depend on the flooding characteristics i.e., the leadtime, water quality, depth, speed, duration of inundation, etc. The hazard is characterized
by its probability and intensity
Hazard (m year-1) = Probability (P) (year-1) × Intensity (I) (m)
Eq. 3-3
Probability is the chance with which a hazard occurs in unit time, which is typically
measured in years in case of floods.
Intensity is the particular characteristic of the hazard that has disastrous consequences.
Floods are distinguished with their exceedance frequency and the probability or average
period that the same intensity flood is expected to reoccur, called the return period
(ADPC, 2005). Generally, frequency analyses are carried out to calculate all potential flood
discharges against their probabilities (Vrijling and Meijer, 1992). The Gumbel distribution
and Log-Pearson type-III distribution are most commonly used frequency curve fitting
methods to estimate river peak discharges. Three types of areas can be marked. One with
frequent flooding where people are highly adapted to floods. In such areas, floods are
welcomed every year, as the livelihood of floodplain inhabitants is mostly associated with
these floods. Second, are the areas that may suffer with only exceptional floods. Such
floods are unexpected and not prepared against, and therefore produce destruction at
mega level. Third, are those areas that suffer flooding after a considerable time. These are
the areas where flood risk is high because of both high probability and exposure. Such
areas must receive most of the flood management considerations.
The intensity can be represented by (a combination of) independent properties of the
hazard. In case of floods, these properties usually are, for instance, water depth, duration,
product of water depth and velocity, etc. For example, the depth and speed are important
to calculate the chances of drowning. A strong current of water (only 60 cm deep) can
carry off most passenger automobiles (APFM, 2007b). The flood depth is mostly
considered an appropriate indicator of flood intensity but other properties are also
considered for many studies. In most cases, the effect of water depth dominates the effect
of other variables (Penning-Rowsell and Fordham, 1994; Wind et al., 1999; Merz et al.,
2007) and thus for the sake of simplicity, most analyses have focused on the relationship
between damages and the stage of flood waters (Yoe, 1994). In Switzerland, the product
of water velocity and water depth is used as a parameter for the flood risk in steep sloping
areas, while water depth alone is used for flat or nearly flat terrains (Meon et al., 2006).
Sediments in floodwater will cause more damages to urban areas, but will prove less
harmful or even beneficial to cultivated land. Therefore, different properties of a hazard
can be used to evaluate risk. The flood intensity at any given location is found by field
data, modeling techniques, or relies on expert judgement. Following is a brief description
of a number of flood characteristics that may be considered relevant in calculating flood
Inundation extent determines which elements are being affected.
Depth is the most prevalent indicator.
Duration of inundation affects crops and determine long-term damages to
Velocity may turn down structures and increases drowning.
Rise rate or warning time lapse is important for flood warning and evacuation.
Time of occurrence during the year is important for crops, and during the day for
rescue and evacuation.
Sediments load or pollutants may increase damages significantly.
The hazard can damage elements that are vulnerable to that specific hazard. If elements
are not vulnerable, there cannot be any risk. For example, if there are no infrastructures
or people in a floodplain, then there is no risk. In a similar way, when people and their
belongings are ideally prepared and protected against flood-damage, then there will be no
risk as well. Thus, the vulnerability is the result of exposure and susceptibility.
Vulnerability ($m-1) = Susceptibility (S) (m-1) × Exposure (E) ($)
Eq. 3-4
Exposure is the value and life that is present within the area under threat (Kron, 2002).
Susceptibility is usually described by relative damage functions (Merz et al., 2007;
Messner et al., 2007).
Exposure is generally related to land-use practices. Land-use maps can be developed by
physical surveys or by consulting GIS maps. These maps show the spatial distribution of
financial, social, and environmental assets. Population distribution, infrastructures, public
services, lifelines, and industrial installations can be identified as important exposures to
be considered while estimating risk.
Susceptibility is the ability to accept the damages from the hazard. The degree of damage
depends upon the susceptibility of vulnerable items and people as well as the intensity of
hazard. This intensity-response relationship (also called damage function) can be
represented using tables, graphs, and equations. Stage-damage (type of intensity-damage)
curves are the essential building blocks upon which flood damage assessments are based
(Smith, 1994). Appropriate damage functions can be derived either from collecting field
data (Genovese, 2006) or synthetic data obtained by lab experiments. Real flood damage
data is collected in many countries directly after a flood has occurred, whereas damage
functions can also be derived from proto-type lab experiments or correlating types of
construction and inventory with different flood properties. The intensity damage
functions behave differently for different land-uses. As mentioned earlier, susceptibility is
the ability to accept the impacts or percentage loss or reduction in value caused by the
immersion by floodwater (Messner et al., 2007). When a flood occurs, it damages
buildings and households, causes drowning of a percentage of the human and animal life
in the inundated area. The extent a building will be damaged, and drowning rates vary
with the capability of buildings and people to withstand these floods. Infrastructures in
floodplains can be built such that these are immune or less vulnerable to flood waters.
Calculation of direct damages
Traditionally, the risk has been defined as the probability times the damage. This
definition is good for calculating risk, but the role of the river and the societal activities,
which cause this risk, remain unclear. In order to understand the role of both sides, it is
necessary to express risk in terms of hazard and vulnerability, as shown in Eq. 3-2. Now,
substituting hazard and vulnerability by their components in Eq. 3-2 we get Eq. 3-5.
Expressing probability separately will result in Eq. 3-6.
Risk = [P × I] (m year-1) × [S × E] ($m-1)
Risk = P × [I × S × E] ($ year-1)
Eq. 3-5
Eq. 3-6
The probability is associated with a hazard that has a specific return period while
intensity, exposure, and susceptibility pertain to the resulting consequences. Hence, Eq.
3-6 represents the elaborated form of Eq. 3-2. This model can be used to the estimate
direct losses. The exposure sets the worst possible damage, while the intensity and
susceptibility determine the extent of this damage, and the probability converts this
damage to the notion of risk.
Figure 3-2: Risk mechanics explained in terms of hazard, vulnerability, probability,
intensity, exposure, susceptibility, and consequences
The purpose of explaining risk in terms of its components is to facilitate the quantitative
analysis of flooding scenarios, while applying different measures. Risk is dependent of all
four parameters. This decomposition of risk into its components is a first simplification or
conceptualization of the more complex reality, which may allow for a better
comprehension and precise analysis of the problem.
Rescinding any of the parameters will nullify the total risk. Numbers of hazards go
unnoticed in everyday life, because of their low risk. There are such examples of hazards
that would disrupt life completely were it not that one of the risk factors makes the risk
negligible. The importance of risk parameters can be illustrated by a number of examples
shown in Table 3-2.
Table 3-2: Daily life hazards with negligible risk
Ultraviolet rays
from the sun
Ultraviolet rays may cause sunburn inflammation
erythema, and nonmelanoma skin cancer, but due to
the low intensity, these are negligible.
Meteoroids may prove lethal to life, but the
probability of hitting human very low.
Dust or pollens are harmful to only those who are
susceptible to them.
Volcano eruptions can be extremely deadly, but
many of such areas do not have close by settlements.
To estimate and minimize the risk of flooding at any location within a floodplain, it is
important to understand the role of basic parameters in inducing risk. The intensity and
probability of hazard, and the susceptibility and exposure relating to the vulnerability are
the basic components that determine the risk. With the understanding of their role in
inducing risk, effective and optimal flood management measures can be selected.
Risk-based flood management practices
Conventionally, flood management practices can be classified as flood abatement, flood
control, flood alleviation, and recovery measures. These practices comprise of structural
and non-structural measures or relate to reducing the activities in a floodplain. Risk-based
assessment also influences the flood management approaches and measures. This section
describes the risk-based assessment of various approaches, plans, and measures, and
highlights the importance of adopting risk-based flood management practices.
Risk-based classification of measures
Selecting effective and most-suitable measures is the most important task in flood
management. Bruijn et al. (2007) has divided direct measures into three types: preventing
flood wave generation (flood abatement); managing the inundation (flood control) and
minimizing the negative impacts (flood alleviation). Sometimes it is important to see
whether the measures involve any structural work that may modify the flow of the river.
Non-structural approaches to flood management comprise those activities which aim to
eliminate or mitigate adverse effects of flooding without the construction of flowmodifying structures (Duivendijk, 1999). Some measures tend to increase the capacity of
society to recover after flooding has occurred. Flood insurance, relief, compensation, and
community support are examples and can be classified as indirect measures, as these do
not reduce losses directly. Relief measures do little to reduce the impact of future flood
losses (Andjelkovic, 2001), and compensation may even increase the risk.
Knowledge of the effectiveness of various measures is essential for designing an effective
strategy. Some measures reduce the hazard by reducing the probability and/or intensity,
where some measures aim at reducing the vulnerability by reducing susceptibility and/or
exposure. Based on risk-mechanics, measures can be classified into two major types; i.e.,
direct measures and indirect measures. While selecting a measure, it is important to note
which target parameter a flood management measure aims at. Different flood
management measures, options and the corresponding risk parameters have been
discussed in Table 3-3. Dividing flood management options based on classification,
structural involvement, and risk-mechanics with target parameters is important to obtain
desirable results for a flood management project.
Table 3-3: Flood management measures and target risk parameters
Target Parameter
Rain harvesting
Soil conservation
Groundwater recharging
Storage and retention
Dikes, Floodwalls
Flow diversion
River re-profiling
River conveyance
Encroachment control
Building codes
Land-use adaptation
Flood proofing
Public awareness
Flood insurance
Exposure, Susceptibility
Susceptibility*, Exposure*
Exposure, Susceptibility
Exposure°, Susceptibility°
Relief efforts
Exposure°, Susceptibility
* Public awareness plays an indirect role
° Play negative role instead, or at least no direct role in reducing risk
Risk-based classification of approaches
Section 3.2.2 describes how a strategy or approach consists of inspirations and priorities
that drive the selection of appropriate measures. The most appropriate strategy will vary
according to the specific geographical, hydrological, social, environmental, and economic
conditions (Green et al., 2000). Risk-based analysis helps to, not only assess and design a
measure (Ref. 3.5.1), but also, choose the most appropriate flood management strategy for
the particular local conditions. Based on the flooding risk-mechanics, flood management
approaches can be divided into two categories
Direct approaches
Hazard management approaches
Vulnerability management approaches
Indirect approaches
Resilience-based approaches
Direct approaches tend to minimize the losses, where indirect approaches target the
prompt recovery from these losses. Direct approaches can be further distinguished as to
those approaches that reduce the hazard and those that reduce the vulnerability of
floodplain inhabitants (Green et al., 2000).
Hazard management approaches follow the principle ‘to keep flood away from people’.
Flood control and flood mitigation are important examples. The approach aims at
reducing the flood event itself by using structural measures. These measures can be
extensive in nature, installed in upper catchments (Tucci, 2007), and tend to reduce the
probability for a flood to occur. Otherwise, these can be intensive measures that are
applied locally (Tucci, 2007) to reduce intensity the flood inundates the area.
Adaptation is an example of vulnerability management (Genovese, 2006) that follows the
principle of ‘living with floods’. The elements of adaptation (societal adjustment) and
preparedness (event-based response (Genovese, 2006)) reduce the susceptibility. In
highly adapted societies, the annual floods are expected and anticipated although the
extreme floods still may cause casualties and damages (Green et al., 2000). Flood zoning
controls the vulnerability by reducing both the exposure (encroachment control) and
susceptibility (flood proofing). Encroachment control is a measure that works on the
principle of ‘keeping people away from the flood’ by reducing the exposure within a
Flood insurance, relief, rehabilitation, and compensation are a number of measures
belonging to an indirect approach. Resilience-based flood management is typical
associated with indirect approaches. The working principles here are to ‘accept floods and
recover afterwards’. De Bruijn (2005) defines resilience as “the ability of a system to
return to its equilibrium after a reaction to a disturbance”. For this, a system must be
designed to aid quick recovery after a flood subsides. Indirect methods reduce the indirect
losses significantly, and indirectly influence the direct losses.
In addition to the above-mentioned strategies, there exist other concepts, like ‘integrated
flood management’, ‘sustainable flood management’, ‘no adverse impact approach’,
‘floodplain restoration’, etc. These approaches are not mutually exclusive and can be
adopted in combination with other approaches. For example, an integrated flood
management prevents isolated perspectives of flood management measures (APFM,
2009a) and does not restrict or exclude any specific measure. Typically, the most
appropriate management strategy will involve a combination of measures and approaches
(Green et al., 2000), and involves complementary options (APFM, 2007d).
Risk-based explanation of plans/ projects
A risk-based assessment enables policy-makers to target risk, and thus results in selecting
the most suitable measures. Risk-based management is geared to the evaluation of
schemes for reducing, but not necessarily eliminating, the flood risk (Pilon et al., 2003). As
mentioned in Section 3.2.4, the severity of a problem and the availability of resources
trigger projects. Furthermore, the type of hazard is also an important factor to set
measures. The response to a hazard depends on the nature of the hazard, i.e., whether it is
natural or man-made and whether the occurrence of the hazard is common. For example,
Nathwani (1997) explained: “Currently, fear of cancer and the risks associated with lowlevel exposures to carcinogenic substances drives much of the regulatory efforts aimed at
minimizing health risks. Diet and smoking, however, cause an estimated two out of three
cancer deaths. They are major causes of cardiovascular disease and deaths”.
The impact of probability on the vulnerability is inversely proportional. Generally
speaking, floods that occur more frequently, are responded to more appropriately and,
hence, cause less damage. Very rare (extreme) floods can produce high damages. These
losses mainly occur in areas that are situated away from the river. The vulnerability
increases, due to more exposure and higher susceptibility as people living at larger
distances from the river, do not expect flooding. The probability of occurrence
predominantly determines the extent of safety measures initiated in a floodplain. The
areas that suffer floods frequently incorporate the measures in infrastructures, like flood
proofing etc. The effectiveness and suitability of flood measures very much depend upon
risk factors as well as technical, financial, and social constraints.
Development of decision-support system
The ultimate purpose of risk analysis is to envisage risk reduction after a proposed project
is implemented and to facilitate decision-makers in selecting the best available option.
Risk evaluation has already been discussed in detail (Ref. 3.4 and 3.5), whereas
communication of results in convenient, elaborated, and guiding shape is discussed in
next sub-sections. A number of different techniques exist to aid in the appraisal process
(e.g. cost-benefit analysis, multi-criteria analysis) (Green et al., 2000). Some basic
concepts, techniques, and terms are discussed here.
Guiding principle for decision-support system
Flood management involves redistribution of risk over temporal and spatial frames. In
our case studies, ‘The Kaldor-Hicks Compensation Principle’ is considered as the prima
facie rule for the decision-support system. The principle states, “A redistribution of risk is
efficient if it enables the gainers to compensate the losers, whether or not they actually do
so”. Flood management measures involve noticeable costs and many of them just divert
floods to less valuable areas. Estimation of benefits while ignoring such losses/ costs are
illusive in nature. Kaldor-Hicks compensation principle emphasizes holistic
considerations of all impacts. The compensation principle establishes the governing rule
that benefits (gains in human well-being) should exceed costs (losses in human wellbeing) for policies and projects to be sanctioned (Pearce et al., 2006).
Structural measures change flood patterns. Such reshaping of the flow regime is
associated with a redistribution of flood risk. As this redistribution reduces risk in highly
populated and industrialized areas, it may cause over-flooding to rural and agricultural
lands at the same time. Net reduction in risk is equal to the combined effects of both areas.
Nevertheless, the protected urban area is still subjected to severe, catastrophic, and less
frequent floods. Structural solutions to deal with this residual risk could be extremely cost
inefficient. Such residual risk could be manageable via insurance at a much lower costs
(Erdlenbruch et al., 2009).
Cost benefit analysis (CBA)
Cost benefit analysis (CBA) is commonly applied to determine the adequacy of a project to
meet its goals. Furthermore, it helps to define the best composition of a project and
compare among competing alternatives (Medina, 2006). The costs include both
maintenance and investment costs of measures. Whereas, in case of floodplain
abandonment, depriving the community from the benefits otherwise it might have by
utilizing floodplain are also considered as costs. Costs are calculated using discounted
rates (Ref. 0) assuming a certain lifetime for the measures. Benefits are damages avoided.
Therefore, the benefits of a project are equal to the difference in damages with and
without project conditions (Medina, 2006).
Cost benefit calculations become difficult if intangible losses and benefits are to be
involved (Kron, 2002). Advanced methods are required to evaluate intangible losses.
Alternatively, CBA can be expanded using multi-criteria framework to consider intangible
along with tangible factors (Tung, 2002) as multi-criteria analysis (MCA) are efficient
tools for the optimization. MCA consider a number of factors to facilitate optimization.
Generally, weights are assigned to different criteria to conclude results of MCA but these
always remain highly controversial and debatable. This approach gives better picture of
the problem (Tung, 2002) but at the same time is unable to compare between different
choices. Therefore, CBA has been selected for our analysis to facilitate decision-makers in
making a choice under the principle of ‘like is compared to like’.
At the same time, summarizing intangible values into a scalar-valued ranking function or
alternatively into monetary terms for the comparison purposes remains debatable until
now (Gilard and Givone, 1997). Establishing a generic valuation of intangible assets, that
might be usable for all types of project at the national level, is highly recommended.
Alternatively, a flexible and elaborated valuation approach to enhance consensus,
reconciliation, and an agreement on valuation of intangibles by separately evaluating
social and environmental losses before combining them into a single EAD value (monetary
terms) can also be used. The general idea of the approach is that the decision-makers
must evaluate intangible losses on fair and uniform basis. A careful valuation of economic,
social, and environmental assets is indispensable to achieve reliable results.
The benefits in CBA are evaluated by comparing damages with and without project
conditions. When damages are compared against different projects, CBA provides a
comparative risk assessment. Comparative risk assessment (CRA) is used to single out the
most effective flood management scheme in context of appropriate flood management
options. The plan with the lowest risk is considered the most appropriate one. CRA
involves analyzing risks for several alternative projects or policies (Pearce et al., 2006).
CRA is used to obtain ORP for designing the master plan for flood measures and for most
suitable land-use planning in floodplain (Figure 3-3). Use of CRA for individual projects
might not produce satisfactory results. Therefore, isolated evaluation of projects without
considering the optimum risk state of the complete strategy might be misleading.
In our case studies, damages due to floods and costs related to measures as well as
benefits are expressed in terms of expected annual damages (EAD) (See 4.3.4). For the
optimization of flood management, multiple reiterations of analysis are carried out under
a number of scenarios in both study areas. Comparing EADs with different measures
helped us to determine ORP in our case studies. In the planning phase, economic and
financial efficiencies of measures may be reassessed to assign priorities to individual
projects under the situational constraints and preferences. Results of this analysis may be
presented in the terms of EAD, IRR, BC ratio, economic rent (ER), and net present value
(NPV) to facilitate investors, decision-makers, and stakeholders to understand the
advantages and limitations of proposed measures.
Economic efficiency indicators
The results of CBA are expressed in terms of economic efficiency indicators. These
indicators help to select the most suitable and the most efficient measure. When several
alternative options are available, the project with the highest economic efficiency could be
the optimal choice. Sometimes, a number of derived indicators are also used for the
decision support system. A brief introduction of some of basic terms is described here.
Benefit-cost ratio (BC ratio)
Benefit-cost ratio expresses the proportion between what is spent and what is achieved. It
is one of the CBA products most commonly used for decision appraisal and is a very
effective indicator of the confidence which can be placed in the decision (Messner et al.,
Internal rate of return (IRR)
Another important term to evaluate the feasibility of a project is the ‘internal rate of
return’ (IRR). IRR provides the interest rate of return of a project through its lifetime of
cash flows. If the NPV is equal to zero, then the IRR is equal to the discount rate. IRR
means the discount rate at which the costs of the project lead to the benefits of the project
(Yi et al., 2010). The World Bank uses IRR as a qualifying indicator for funding the
Figure 3-3: Comparative risk analysis of different strategies and calculation of optimal risk
against different land-use schemes in floodplain
Economic rent (ER)
Reduction in flood losses to existing land-use can also be expressed in terms of increase in
land-use efficiency. Economic rent (ER) helps in assessing the land-use efficiency.
Economic rent is commonly defined as the net annual income associated with a resource
(Weisz and Day, 1975). ER of a land-use may be computed as the average of the annual net
returns discounted to their present value. If ‘Rn’ is the annual net return from unit land
area in year ‘n’, assuming a constant discount rate ‘r’, then the economic rent ‘ER’ over the
time ‘t’ will be the average of discounted annual net returns. This relationship can be
expressed in the following way (Eq. 3-7):
𝐸𝑅 = �
𝑡(1 + 𝑟)�
Eq. 3-7
The net annual return to unit land can also be defined as the gross annual return minus
the annual total non-land costs and is synonymous with economic efficiency returns
(Weisz and Day, 1975). In case of floodplain, annual costs on measures and remaining
expected damages reduce economic rent. This combined reduction may be terms as
‘combined deductions’ (Figure 3-4). In terms of economic rent, ORP can also be defined as
a state of flood management at which risk deductions to economic rent are minimal.
Economic rent might be the most suitable index to be used to evaluate the combined
effectiveness of more alternative means of attaining floodplain management objectives. It
has two components, namely, ‘location benefits’ (Rloc) and ‘intensification benefits’ (Rint).
Location benefits can be defined as the value of making floodplain land available for new
economic uses, such as shifting from agricultural to industrial use (USACE, 1996).
Intensification benefits are the value of intensifying use of the land, such as shifting from
lower to higher-value or higher-yield crops (USACE, 1996).
∆𝑅� = ∆𝑅�𝑜� + ∆𝑅𝑖�𝑡
Eq. 3-8
As flood management is improved, both components rise with higher flow of investments
due to reductions in negative impacts of floods. Although increase in location benefits and
intensification benefits depend on a number of factors, availability of investments play a
main role. Precise evaluation of economic rent is subject to the planning of investment
projects in the floodplain. Location benefits are considered while implementing the flood
zoning in our case studies (See 6.5).
Figure 3-4: Combined deductions and minimum required net-benefit of land-use
Discounting procedures
Although floods cause damages to life and infrastructures, flood measures also need heavy
investments to reduce these damages. Investments in flood measures are ideally low and
are paid in advance whereas flood damages are high and might occur in the future with
some probability. Flood measures provide benefits throughout their useful life by
reducing the losses, equal to difference in EAD, every year. Flood measures need capital
investment in the start and some maintenance costs annually. This capital investments
may produce some annual return (equal to compound interest or discount rate ‘i’) if these
were deposited in a bank. Therefore, future returns from flood measures cannot be
considered equal if these were obtained at present time. The concept of present value
(PV) handles these two series of unevenly distributed benefits and costs.
PV is a basic concept of economics that accounts for the time value of money. Discount
rate is considered in all analysis that compares investments and outcomes. The discount
rate used is typically set by the federal government and is much higher in developing
countries when compared to developed countries. Present value ‘PV’ of investment ‘P’
with ‘i’ interest rate over time ‘t’ (in years) can be found using Eq. 3-9.
𝑃𝑉 = 𝑃�(1 + 𝑖)𝑡
Eq. 3-9
In our case studies, interest rate was kept uniform at 12%. In a CBA, it is usually
preferable to present the costs and benefits difference in terms of present values to be
able to compare costs and benefits from different (future) timestamps. Once the PVs of
benefits and costs have been estimated, the BC ratio and net benefit (NB) of the project
can be computed using the formula Eq. 3-10 and Eq. 3-11 (Pearce et al., 2006):
𝐵𝐶 𝑟𝑎𝑡𝑖𝑜 =
Eq. 3-10
𝑁𝐵 = 𝑃𝑉� − 𝑃𝑉�
Eq. 3-11
𝑃𝑉� > 𝑃𝑉�
Eq. 3-12
If the net benefits are positive, then the project is cost-efficient and the BC ratio is greater
than one. The necessary condition for the adoption of a project is that discounted benefits
‘PVb’ should exceed discounted costs ‘PVc’ (Eq. 3-12) or if net present value ‘NPV’ is
greater than zero (Eq. 3-13).
𝑁𝑃𝑉 > 0
Eq. 3-13
As mentioned, any project for which benefits exceed costs can be accepted, however, if
resources are limited, then not all ‘acceptable’ projects can be undertaken. In this case,
projects must be ranked or ordered in terms of the objective function and available
resources. Priority ranking of projects is always based on BC ratios. On the other hand,
when adequate resources are available (and ideally should be available according to the
stipulations behind the risk-based approach), the project with the highest net benefits
(not the highest BC ratio) is the optimal choice (Medina, 2006). Same is recommended for
both the proposed flood management standards and flood management optimization in
our case-studies. A modern concept that is still evolving in many countries points at
maximizing the net-benefits from floodplains, rather than aiming solely at minimizing
flood damages (APFM, 2007b). This idea is further pursued in case studies (See 6.5) and a
concept of ‘floodplain management’ has been recommended (See 8.3).
Optimal risk point (ORP)
As mentioned in sub-section 3.3.1, the purpose of flood management must be reducing
risk. Risk cannot entirely be nullified in such situations. It is also mentioned that, flood
measures reduce flood damages but involve high initial investments and operative
maintenance costs. Therefore, the risk in a floodplain cannot be lowered beyond a certain
threshold. At this threshold, the combined annual expected losses due to flood damages
and costs of measures are at the minimum.
In a floodplain, measures taken in the initial stages usually have high BC ratios. BC ratio
then decrease for measures taken afterwards, and may become one or even less for
subsequent measures. The stage where marginal costs on measures and marginal benefits
are equal, expresses the ‘optimum state’ (Figure 3-5) (Park, 1999; Jonkman et al., 2009).
Different measures may demonstrate different optimum states. The optimum state with
lowest risk under all favorable combinations of flood measures is known as ‘optimal risk
point’ (ORP). Figure 3-6 explains the concept of ORP when different strategies (consisting
of suitable measures) are compared. It also compares the feasibility of a proposed landuse practice. It also supports the idea that maximum land-use benefits can only be
obtained when ORP is achieved in floodplain.
Figure 3-5: Relationship between increasing costs on measures and reducing flood damages
Figure 3-6: Comparative risk analysis of different strategies and calculation of optimal risk point
against a land-use scheme in floodplain
The strongest point of probability-based approach is that these provide countrywide
uniform standards of practices. Risk-based approaches so far fail to provide standards
that might uniformly be practiced over a country. With the help of ORP, flood
management may be described in terms of risk-based standards. The current or proposed
status of flood management might be described as the ratio or percentage of ORP in a
floodplain. This ratio may be termed as ‘flood management ratio’ (FM ratio). FM ratio
expresses the ratio between present EAD and EAD at ORP. Ideally, the FM ratio in a
floodplain should be unity unless there is temporary shortage of resources and achieving
ORP is planned stepwise in phases. Additional information may also be provided by
expressing the ratio between the losses due to floods and the investments on the
measures for planned EAD.
The commonly practiced economic efficiency indicators need to be used carefully for the
assessment of flood management plans. There are two broad categories of these
indicators. Some indices compare input-output ratios and the remaining compare inputoutput differences (Eq. 3-10 and Eq. 3-11). BC ratio is based on input-output ratio. Costbenefit difference, ER, IRR, NPV, and EAD are based on the difference (or difference rate)
of inputs and outputs. These indicators (except EAD) are designed to evaluate the benefits
of an investment and can be used for evaluation of flood measures investments as well.
Nevertheless, these indices should not be used in a straightforward way for flood
management planning. Such practices might be misleading and may derail the planning
strategies away from achieving the ORP. The maximum possible ER of a floodplain with
available resources is only possible at ORP (Figure 3-4). The most efficient master plan
should be decided using proposed ORP standards. For the situations, where there is a
budget constraint, the priorities to individual measures within the master plan can be
assigned using established indicators (mostly BC ratio) suitable to the situation and
Figure 3-7: Flood management costs, benefits, BC ratio, NPV, and ORP concepts
Rivers are the source of life and livelihood by enabling fisheries and providing fresh water
for domestic, industrial, and agricultural usage. Even floods have numerous advantages
along with disadvantages. Risk-based assessment considers these advantages of land-use
and river proximity, while accounting for the potential losses. That is why, risk-based
assessment is considered a tool to balance river processes and societal activities (Ref.
3.3.6). Risk-based flood management has the ability to handle floodplain dynamics and is
inherently capable of achieving an (dynamic) equilibrium between human activities and
river processes (Nathwani et al., 1997). In particular, for developing countries, where
vulnerability is changing quickly, due to their fast economic transformation (Benson and
Clay, 2004).
Another main advantage of risk-based flood management is that it helps to select efficient
measures according to the nature of the problem, because of its ability to minimize the
disaster impact (Genovese, 2006). In fact, risk-based assessment means to evaluate the
impacts instead of controlling the floods. It prioritizes measures that can address the
problem in effective and efficient way. Employing risk-based assessment is therefore the
only option to optimize flood management.
A risk-reduction approach for organizing flood management plans likely would produce
the desired results given its ability to consider all possible options. With its flexibility to
work with almost all major in-practice approaches, it provides a broad canvas to examine
various different measures. To conclude, it is agreed that the risk-based approach should
be adopted at all levels of flood planning (DCLG, 2006). Thereby, recognizing the necessity
to move towards a risk-based approach, the European Parliament has set an example
(Moel et al., 2009).
It is evident that establishing ORP is the first step towards designing a flood management
strategy. However, FM ratio may be used to plan stepwise planning to achieve ORP at
national level. FM ratio also helps by indicating the current departure from achieving ORP.
Whereas IRR, NPV, and benefit cost ratio are indicators that describe the feasibility of any
measure or project within the master plan. Preliminary optimizations of both the land-use
and flood measures have been performed in this thesis.
The maximum land-use benefits in a floodplain are achievable at ORP. Therefore, the first
step towards the planning of flood management is to determine ORP of a floodplain. Long
term and short term planning may be formulated consisting of only those projects that fit
in the strategy that achieves the ORP. Therefore, priorities may be assigned only to those
projects using BC ratio (Ref. 3.6.2). However, extensive efforts for data processing and
analyses are required to establish the ORP. The proposed risk-based standards, if
developed further may replace probability-based standards. ORP once established must
be re-evaluated on a periodic basis.
4 Flood management assessment
This chapter will explain the recommended methodology for implementing the proposed
risk-based assessment (Ref. 3.4) in flood management projects. The framework suggested
here, essentially comprises of both the evaluation of river processes and societal
responses. A flood measure can be considered as an intrusion into the system while such
intrusion may have its impacts, both positive and negative. For example, introducing a
dike may reduce probability of flood in one area and may increase it in another area. The
assessment methodology in such case must cover the impacts on both areas. The method
recommended in this chapter, emphasizes the identification of all involved impacts and
areas. The recommended framework involves probability-based 2D analysis of flood
impacts to describe the temporal and spatial situations due to all possible floods.
The main steps recommended hereby are analyzing the behavior of flood and society,
short-listing the feasible flood management options, evaluation of measures individually
or in combinations, and finally to recommend the best possible scheme. Schematic layout
of these steps is demonstrated in Figure 4-1.
River behavior (Hazard parameters)
Fluvial floods are mostly caused by the fluctuations in river flow. Therefore, their relative
intensity, duration, and recurrence pattern play a major role to shape flood impact.
Average flow of the river is not that much important. With regular fluctuation patterns,
societies are mostly well-adapted and are used to get maximum benefits out of the floods.
River behavior perhaps is the most important factor to be considered while analyzing the
flooding problem. It is the important factor used for designing measures according to
probability-based design.
Figure 4-1: Proposed schematic layout of risk-based assessment
Flood frequency analysis
Flood risk is a function of damages and their probabilities. The occurrence of flood and
resulting losses, both are probabilistic. The probability in occurrence of flood is estimated
by frequency analysis, whereas, probability of losses occurrence is generally treated while
developing intensity-damage functions (Ref. 3.4.3). Frequency analysis provides a
probabilistic approach to design measures assuming that events in the future are
predictable based on the experience of the past (Pistrika and Tsakiris, 2007). The purpose
of flood frequency analysis is to relate the peak flows to their frequency of occurrence
using probability distribution functions (Chow et al., 1988). Fitting the annual peak-flows
to the most closely matching cumulative probability distribution function is the most
straightforward method for estimating flooding probabilities (Francés and Botero, 2007).
Flood frequency analyses are traditionally practiced to deal with flood probabilities. Main
limit is that these assume that annual maximum floods are stationary, independent,
identically distributed random processes.
Frequency analysis has been carried out to estimate the probability of potential peak
discharges for the Swan (at Islamabad Highway Bridge) and Chenab (at Marala
Headworks) rivers. In our case studies, preliminary tests for independence and
homogeneity as well as tests for outliers are carried out. The best-fit distribution was
identified based on Chi square value using the L-moment ratio diagram, best fitting visual
inspection, and the Z-dist statistic criteria. ‘Design Flood’ software developed by Iftikhar
Ahmad (Ahmad, 1994) is used to perform frequency analysis and calculations are double
cross checked by IH-Flood and Xtremes 4.1 developed by Centre for Ecology & Hydrology
and Xtremes Groups, respectively.
Annual peak flow data of 83 years are available in case of the Chenab River study area.
Frequency distribution based on collected data follows the Log Pearson Type III
distribution, calculated with method of moments. Ranging from 0.0001 to 0.5
probabilities, ten different flood scenarios are estimated. Floods for 0.0001, 0.0002, 0.001,
0.002, 0.01, 0.02, 0.04, 0.1, 0.2, and 0.5 probabilities are calculated. Flow measuring gage
of the Swan River is much below the study area and flow data for 41 years and rainfall
data for 32 years are available. Rainfall-runoff modeling is performed to estimate peak
flows precisely. Frequency analysis results of both rivers are shown in Figure 4-2.
Figure 4-2: Flood frequency analysis for both study areas: the Chenab River at Marala Headworks using Log Pearson Type-III distribution, [MM method], the
Swan River flows at Dhok Pathan and annual maximum rainfall at Islamabad Airport using Generalized Extreme value Distribution, [ML method]
Rainfall-runoff modeling for Swan study area
The hydrological modeling of the rainfall-runoff process is performed for the Swan River
study area using HEC-HMS software developed by US Army Corps of Engineers (Army
Corps of Engineers, 2011). The catchment of the study area is subdivided into 39 subcatchments on the basis of the available geographical information. The minimum area
limit for subdivided catchment was kept at 4km2. These cover the whole watershed area
and represent drainage units in which the hillslope runoff (surface and subsurface)
converge to one point, the drainage point of the each sub-catchment. All sub-catchments
are linked by channel sections (Figure 4-3). Their ends match with the drainage point of
the sub-catchment and drain into the adjacent channel section. Still in each sub-catchment
the vegetation cover, land-use and the hydrogeology might vary. The subdivision of the
catchment is based on topography retrieved from a Digital Elevation Model (DEM) of the
catchment with 25m resolution (Figure 4-4). DEM data is obtained from ASTER website
(ASTER, 2011).
Figure 4-3 shows the schematization of the catchment and illustrates the connections
between the sub-catchments. The rainfall-runoff transformation for each sub-catchment
was modeled. Losses were deducted and the flood wave was obtained at the mouth of
catchment using hydrologic routing along with base-flow. Rainfall to runoff conversion
depends on land-use. For our case study, land-use data was obtained by Cheema and
Bastiaanssen (2010) (Ref. Figure 4-4). Appropriate methods/ models within the package
were chosen depending on the available data and the amount of accuracy required.
Processes and methods selected in our case study are described in Table 4-1.
Table 4-1: Details of methods selected while modeling rainfallrunoff using HEC-HMS
Method Selected
Rainfall distribution
Canopy losses
Evenly distributed, reduced by 20%
Not incorporated
Initial and constant
Clark unit hydrograph
Not incorporated
Figure 4-3: Schematic layout of HEC-HMS rainfall-runoff model of the Swan River catchment
Figure 4-4: Land-use at 1km resolution on left (Source: (Cheema and Bastiaanssen, 2010)) and
right side figure represents the hydro-topographical features of the Swan River study area
catchment at 25m resolution.
Flood simulations
The flood behavior and societal response, both together, determine the extent of
consequences. The nature of the flood hazard must be modeled, before designing a flood
measure (Green et al., 2000). Nowadays, models of different complexities are applied for
the hydrodynamic simulations of floods to analyze flooding characteristics. Remote
sensing/ GIS data are becoming more commonly used for simulating floods by modern
software packages as these data are cost efficient compared to surveys (Hussain et al.,
2009). ‘Sobek Rural’ developed by ‘Deltares’ is used for 1D-2D simulation of flood. The
Sobek 1D-2D module performs 1D modeling for Channel flow and 2D for the flood-spread
over the floodplain. For complex geometries and flow patterns, 2D modeling is
recommended (Kron, 2007; Forster, 2008). However, the 2D model approach requires a
significantly higher processing effort and longer computation times (Forster, 2008). In
our case studies, one-dimensional models of the river coupled with 2D overland flow
using ‘digital elevation model’ (DEM) were used to produce desired results.
The simulation of inundation depths and floodplain extents for the Chenab River area was
carried out using 90m resolution DEM obtained from the HydroSHED (Lehner et al., 2008)
in Sobek Rural ‘1D-2D’ module. The model has been run against a selected range of floods
frequencies (Ref. 4.1.1). Schematic layout and hydraulic simulation was done using HEC
RAS, HEC GeoRAS, and Sobek ‘1D-2D Rural’ modules. The model has been run for all
calculated floods frequencies. The study area is flat and may exhibit low velocities and
longer duration of flooding. Floodwater depth was selected assuming the adequate
parameter for loss calculation. The Chenab River model was carefully calibrated and
validated against available data of water levels recorded at the Marala Headworks and
Qadarabad Headworks. Initial Manning’s roughness value ‘n’ were chosen following the
literature and models developed by NESPak Consultants. The Swan River is smaller than
the Chenab River. Therefore, the DEM of 25m resolution is used. There is no flow-gauging
site in the study area that can be used to validate our model results. Due to lack of
calibration data for the Swan River, ‘n’ values obtained from the literature review were
used. This assumption might be acceptable, since there is no other possible method, not
even with available GIS data, to calibrate our model.
Societal response (Vulnerability parameters)
A flood of the same intensity may produce different losses in different areas due to the
difference in exposure as well as resistance and preparedness of the societies. Flood
response of society is often characterized using intensity-damage curves. For this
purpose, we used the function developed for areas similar to our study areas after
consultation of experts. Inundation of the floodplain was simulated using 2D flood
propagation model (Sobek) for a number of inundation scenarios with various measures.
These scenarios were developed using dikes of different heights and combinations of
land-use practices. Land-use practices and preparedness determine the societal response.
Land-use practice (Exposure)
Land-use maps for the Chenab River floodplain were developed using ‘Pakistan survey
department’ maps and were calibrated with ‘google maps’ and partial ground-truthing
techniques. Land-use maps of the Swan River were prepared by combining the
development plans of land-developers currently developing the floodplain in the study
area. Figure 4-5 and Figure 4-6 show our study areas of the Chenab and Swan rivers.
These maps were used to display spatial distribution of exposure.
Flood response (Susceptibility)
Flood response (Susceptibility) is the property that depends upon the ability of assets or
people living in floodplain to withstand against floods (Ref. 3.4.3). As mentioned in
subsection 3.4.3, susceptibility of a land-use can be defined using intensity-damage
relationships. The stage-damage functions are accepted as the standard approach to
assess direct flood damage (Smith, 1994; Middelmann-Fernandes, 2010; Yi et al., 2010)
and the same is used in our research. Accurate information on historical flood losses are
very rarely available; therefore, estimates are often accompanied by large margins
(Douben, 2006). In our case, appropriate damage functions have been developed by
consulting literature, different sources, and previous projects on the best-suited basis. Our
dominant sources were local authorities, Chen 1999, Wang & Xiang, and ANFAS project
(Data Fusion for Flood Analysis and Decision Support). Figure 4-7 shows the stagedamage functions ‘Fa (d)’ used in our case studies. If ‘Da,max’ is maximum possible damage
for a particular land-use ‘a’, then damage ‘Da,i’ against a flood of probability ‘i’ can be found
using Eq. 4-1.
Da ,i =Fa (d ) × Da ,max
Eq. 4-1
Risk assessment
Once the flooding behavior and societal response is determined, the next step is to
calculate the risk (potential consequences) with and without project conditions. Choosing
a standard procedure, level of details, types of damages, and proper software tools are
discussed in next sub-sections. The ultimate outcome of proposed assessment
methodology is the detailed estimate of risk. The convenient presentation of the results
facilitates decision-makers to choose the best possible option.
Flood damage assessment includes both engineering and economic aspects (Yi et al.,
2010) and can be performed with a number of techniques. The choice of an appropriate
method depends on too many different factors and not least on the data availability.
Almost none of developing countries have any standard methodology for loss estimation
(Dutta et al., 2001). However, there exist some generally accepted methods for the loss
estimation. Depth-damage functions described in subsection 4.2.2 are used to estimate
direct damages.
Figure 4-5: Chenab River floodplain showing the digitized land-use details.
Figure 4-6: The Swan River floodplain showing the developed areas by different developers.
Figure 4-7: Depth-damage function used for case studies
To help decision-makers, there is a need as far as possible to devise a common metric for
all the consequences of flooding whether tangibles or intangibles (Schanze et al., 2006).
Efforts are made to include all possible types of damage categories that include direct,
indirect, induced, tangible, and intangible damages to consider the net impact of flooding.
Indirect tangible and intangible damages may constitute a significant contribution to total
damages. Unlike the estimation of direct losses, the modeling of indirect loss is still in a
preliminary phase. Simple, precise, accurate, and standard methods to estimate indirect
and intangible losses are still far from maturation (White et al., 2001; Veerbeek, 2007).
Current methods of flood loss estimation, either retrospective (i.e. backward analysis
based on empirical data) or prospective analysis (i.e. scenario analysis and prediction),
mainly focuses on direct losses (Veerbeek, 2007). In this thesis, direct and tangible
damages are calculated in detail, whereas, indirect and intangible damages are estimated
using generally practiced correlations that can produce fair estimates (Yoe, 1994). In
practice, only important damage categories are used in different case studies depending
upon their potential role to keep the effort of the analysis reasonable. In addition to losses,
floods have many benefits. Flood inhabitants also have interests that are sometimes
associated to river flooding. Benefits of floods and living in river proximity are also
considered while calculating risk in our case studies.
Scale of the study
Taking a decision about the size of study area is the first step before the execution of a
flood damage-evaluation study. The selection of damage evaluation method depends very
much although not entirely on the availability of data as well as on scale and objective of
the studies. Scale of the studies is linked with the type and required details of the
assessment. The choice of an appropriate type of damage assessment method depends
very much on the size of the study area (Messner et al., 2007). If a method of damage
evaluation involves more details then it will need consequently more efforts per unit of
area. Consequently, the type of ‘monetary analysis’ is also related to the scale of studies.
For small-scale studies, damages to individual household, business plant, or company are
considered as ‘financial damages’, while the macroeconomic effects at country or regional
level are covered under ‘economic damages’ (De Bruijn, 2005). “The ‘financial analysis’ of
a project estimates the profit accruing to the project-operating entity or to the project
participants, whereas ‘economic analysis’ measures the effect of the project on the
national economy” (ADB, 1997). The size of study area and detail-level of analysis must be
decided based on available resources and objectives of study.
Dwelling household
Dwelling building
Electric line
Sand dune
Barren land
Major road
Secondary road
Main canal
Tributary canal
Max. %
Land use
Damage Depth
Sr. Nö
Table 4-2: Land-use classification used in our case studies for calculation of direst
tangible losses
Total Damageable
Tangible losses
Direct tangible damages are estimated using the damage functions (Figure 4-7). These
damage functions estimate losses as percentage of the exposure of each asset type. The
percent damage, thus obtained, is then multiplied by the exposure to obtain damages in
monetary terms. The depth–damage functions are used to estimate flood damage by using
flood inundation depth map and exposure distribution map (Ref. 4.2.2). Intensity-damage
function of one asset differs from others. Some assets are movable, may be moved away
from flood risk areas, given sufficient warning time, and are not necessarily potentially at
risk. Therefore, a percentage is assumed to be flooded depending upon warning time and
immovability of assets. Land-use classification, effective flood depth, and maximum
percentage of damageable costs are shown in Table 4-2. Thus, it is obvious that damage
functions count the fact that floods do not results in total collapse. In addition, some
property can be removed from floodplain at the event of flooding. These functions also
consider the probability that a loss will occur each time under same conditions.
Indirect Losses
Due to the involved complexity in estimating indirect losses, much less attempts have
been done to design such models (Veerbeek, 2007). Indirect losses are mostly correlated
to direct losses, or otherwise, treated in a very complicated way. Unit-loss models, inputoutput models, econometric models and fixed ratio models are used mostly for the
estimation of indirect losses. It may be acceptable to use a fixed ratio between direct and
indirect flood losses in some studies (Yoe, 1994). The ratio of indirect to direct national
losses is 12.7% at 1 meter flood depths (Yoe, 1994) and the same is used in our case
studies where needed. Care should be taken while estimating indirect losses to avoid
double counting of damages.
Intangible losses
Actual worth of any asset is the sum of its tangible and intangible values. The proportions
between both values vary greatly according to type of asset. Intangible values of
commercial products are almost negligible, whereas, of historical monuments, religious
places, wetlands, habitat of rare species, and graveyards are higher than their tangible
values. Households are consisting of some items of sentimental or nostalgic value for
which market values that only covers tangible value, may not measure their true worth.
Intangible losses can be of either social or environmental in nature. Environmental and
social values have been evaluated into monitory terms for our cost-benefit analysis using
commonly practiced methods. Because such evaluation presents considerable challenges
and difficult choices (Schanze et al., 2006), the framework in our research is kept flexible.
Social and environmental can also be described explicitly before concluding a final EAD.
Estimation of life losses
Loss of life and affected population comprise the most common and dominant fraction of
social losses. Number of persons affected can be estimated by averaging the population
over the area while correlating the life losses to the flood parameters. Roos and Jonkman
(2006) have distinguished three categories of flood deaths
Drowning persons due to rapidly rising water
Drowning persons due to high flow velocities
Deaths due to other causes, such as hypothermia, heart attacks, shock, failed
rescue, etc.
It is expected that especially the combination of water depth and rate of rising cause the
danger. A function was developed (Eq. 4-2) that based on the fact that mortality ‘K’ is a
function of water depth ‘h’.
Eq. 4-2
K = f (h) = 9.18 ×10−4 e1.52h  f (h) ≤ 1
Although potential loss of life is noted in many feasibility reports, there are no welldefined regulations or guidelines for how to incorporate loss of life into the cost-benefit
analysis. A more general relationship has been described by Pelling et al. (2004). He
described the number of causalities ‘K’ in terms of exposure ‘E’, gross domestic product
per capita ‘G’, and local population density ‘D’.
ln(𝐾) = 0.78 ln(𝐸) + 0.45 ln(𝐺) − 0.15 ln(𝐷) − 5.22
Eq. 4-3
Eq. 4-3 is used in our case studies. These relationships are not useable for social hot spots
like hospitals, schools, old people’s homes etc.
Expected annual damages (EAD)
Risk-based methods of evaluation help decision-makers to envisage spatial and temporal
distribution of risk over the floodplain. Flood damages are probabilistic events (Yoe,
1994) and flood intensity greatly varies over the floodplain. The spatial distribution of
risks as well as of the benefits of flood mitigation measures have been rarely considered
so far (Meyer et al., 2009). Expected annual damages (EAD) represent the average
damages that might be expected annually due to the probabilistic nature of floods and
flood losses. Damage curves and EAD distribution maps provide an expanded illustration
of risk distributions. EAD can also be calculated across the flow to see the lateral trends in
risk along the river. Therefore, decisions, which are based on EAD, have advantages over
conventional approaches. EAD provides the estimation of risk and indicates how far
present flood management practice is from achieving the optimal risk point ‘ORP’.
In 1990’s, US Army Corps of Engineers used the relationships between the damage,
frequency, discharge, and stage to develop a hydroeconomic model for EAD estimation
(Yoe, 1994). These correlations work well as long as flow conditions and societal
vulnerability remain unchanged. Structural measures change the flow regime and flood
behavior significantly. Once the flow regime is changed, old developed relationships do
not remain applicable. Therefore, a hydroeconomic model cannot be used in such
conditions. Detailed analysis using GIS data and 2D hydrodynamic model are needed to
handle such changes appropriately. The method supports the calculation of EAD based on
actual conditions of the floodplain. The concepts of EAD, damage curve, and EAD
distribution map have been introduced along with their possible benefits and proposed
EAD = � D� × ∆P�
Eq. 4-4
𝐷𝑖 =
𝐷𝑃𝑖−1 +𝐷𝑃𝑖
Eq. 4-5
The term ‘expected’ is used rather than ‘average’ because a frequency curve is used to
represent the distribution of future flood events and the expected value of damage is
computed by the summation of probability weighted estimates of damage (USACE, 1993).
The unit area (Yardstick) approach
The Yardstick approach is an effective method for small-scale detailed studies. This
approach considers individual properties and assesses damage per square meter of the
floor space (Messner et al., 2007). The price of any land-use type is determined based on
involved materials and labor or otherwise by consulting state agencies, experts, and
property brokers to establish market price per unit area. In our case studies, the latter
method is used to establish the price of different land-uses and is crosschecked with the
first method. Unit loss models require detailed information of land-use. The price of landuse (excluding the price of land) expresses flood damage potential. Flood damage
functions calculate losses as percentage of these potential damages. Whereas, the
percentage is calculated using the relationship between flood intensity parameter and
land-use susceptibility. Generally, depth-damage relationships are used for damage
calculations (also in our case studies, Figure 4-7) to find the percentage of land-use value.
This method can also be used to calculate the casualties and population at risk (De Bruijn,
Another advantage of this approach is that it may provide the density distribution of
damages over the floodplain. The damage density distribution helps in pointing out the
high vulnerability spots. The resulting damage maps are useful for interactive flood
management. This method is helpful to translate these damages in terms of risk.
Software tools
A number of software programs have been developed for flood damages calculations.
Calculation of flood damages using a complete software package reduces the efforts
considerably. At the same time, it limits the freedom of incorporating diverse and new
methods for research purposes. These methods can mainly be divided according to their
ability to work with GIS data. There seems to be a trend that as GIS has become more
common over the last years, it is being integrated in damage evaluation software tools
more often (Messner et al., 2007). Software tools also differ in their capabilities and
methodologies. Some are capable to calculate flood risk from the basic hydrologic,
hydraulic, and topographic inputs while others only calculate flood damages against flood
intensity and land-use parameters (Van Mierlo et al., 2003). Table 4-3 demonstrates some
popular software tools with their GIS capability and countries of origin.
Table 4-3: Popular software tools, their compatibility to GIS and their country of origin
Czech Republic
South Africa
South Africa
Uncertainty and risk analysis
An important aspect of risk analysis is that the input information for the analysis can
never be perfect and methods employed approximate real life. Therefore, outcomes of
such analysis can be described as an educated guess of reality and lack certainty. In
scientific terms, ‘Uncertainty’ is the difference between the statistics of the sample and the
population of a set of data (Tucci, 2007). Flood-estimation procedures in practice mainly
rely on observed discharge data (Buchele et al., 2006) and the quality of source data used
in analysis brings major uncertainty. Another key aspect in modeling uncertainty depends
upon the availability of actual observed flood events used for the model calibration
(ASFPM, 2004). The flood simulations are based on ‘clear-water’ flooding which happens
seldom in reality (Myers, 1995). Debris blockages often create flooding conditions in
unexpected areas. The vulnerability of elements at risk is also a source of uncertainty
(Merz et al., 2007). Uncertainties may also arise as a result of the land-use data, the value
assessment, and mostly in the estimation of susceptibility, i.e. the damage functions
(Messner et al., 2007).
Uncertainty plays a major role in the credibility of results. Decision-makers must be aware
of the reliability of the risk analysis results while assigning priorities and resources to
projects. It is important to consider all the uncertainties to obtain accurate estimate of
EAD; otherwise, EAD would be underestimated mostly (Tung, 2002). Mentioning
uncertainty in results enhances the understanding of decision-makers. Risk analysis
results should, as far as possible, indicate the error range and provide the user with a
realistic idea of their accuracy.
All uncertainties cannot be avoided or solved. Choices are, by necessity, made under
uncertainty; because they lie in the future, the consequences of all the options are
shrouded in uncertainty (Green et al., 2000; De Bruijn, 2005). Most important to consider
is to foresee the impacts of such decisions. Decisions under high uncertainty must be
adaptive and flexible and need much effort to minimize the uncertainty. Considerable
uncertainties, although being an intrinsic part of extreme value estimations, can be
managed by the complementary use of different methods (e.g., flood-frequency analysis
and rainfall-runoff models) (Buchele et al., 2006). Schanze et al. (2006) recommend the
use of a probability distribution from Monte-Carlo modeling or from more simple
representations as ‘credible’ bounds on some key parameters to represent and
communicate the uncertainty in the assessments. To decide whether an uncertainty
analysis is required depends upon the extent and reliability of input information available
and purpose of the study. On the other hand, a very approximate damage estimate might
be sufficient for a high level strategy decision but not for the assessment of concrete
protection measures (Messner et al., 2007).
Design optimization
According to the proposed framework (Figure 4-1), the last step is the optimization of the
design. Our framework provides the results of risk estimation in shape of EAD values,
damage curves, and EAD distribution maps. These values help the designer to modify the
proposed measures interactively to further reduce risk.
The EAD distribution map provides the most detailed spatial distribution of risk. This
might be helpful to diagnose the high-risk areas in floodplain. These maps provide the
combined risk due to all possible floods and cannot show the impacts of individual floods.
Damage curves facilitate to see the impacts of any individual flood. If the corresponding
damages are weighed according to their (exceedance) probability, these curves represent
the EAD curves that demonstrate the contribution of individual flood to risk. The lump
sum and net impact of any measure is described by its EAD value.
Figure 4-8: Comparative risk analysis of different strategies and calculation of optimal risk point
against a land-use scheme in floodplain (Reproduced).
Every modification in design of a measure will change the EAD of the floodplain. If this
EAD is drawn against the costs required for the measures, a trend line similar to Figure
4-8 can be obtained. Every measure or combination of measures will reduce risk with
increasing costs. These ‘combined deductions’ (Ref. 3.6.3) continue to decrease and then
start to increase after reaching an ‘optimal state’. A comparison of optimal states against
different measures or combination of measures provides the ORP that should have the
minimum combined deductions.
We have evaluated the optimal states of dikes, flood zoning, and insurance for our study
areas in Chapters 5, 6, and 7. As the flood frequency remains unchanged in all cases, these
chapters will cover flood simulations, societal response, risk assessments, and optimal
states of proposed measures.
The proposed framework defines a risk-based assessment and design approach. This
approach is equally applicable for developing and developed countries. The assessment
methodology considers the vulnerability of the society in addition to river behavior. To
design a risk-based flood measure, important general steps have been described in this
chapter. Details for few types of measures (structural or non-structural) are provided in
next chapters for example cases.
Although, the development of probability-based standards are supposed to be developed
according to the socioeconomic conditions of a country, countries adopt these standards
considering standards in other countries (Smith, 2004). Whether a design standard in a
country is developed properly or not, probability-based design methodology only covers
the hazard behavior, usually done by frequency analysis. The risk-based approach
stresses the need to evaluate hazard and vulnerability of society for every single design.
This implies that according to the risk-based approach, socioeconomic conditions of
floodplains cannot be generalized and must be evaluated on a case-to-case basis. The
proposed framework describes the steps that are required for the detailed analysis.
Concluding, this chapter provides the general considerations that are required to address
the flood measure design under different conditions. It addresses the necessary steps to
define the hazard characteristics and societal response. It also covers the level of detail,
types of losses to be considered, damage calculation methods, and an overview of
available software packages.
5 Hazard adjustments
The first obvious option to reduce hazard risk is to control the hazard itself. In the field of
flood management, hazard control has been the most popular and economically feasible
option most of the times. For example, fluvial flooding in flat areas inundates wide areas
around the rivers. If a literal application of ‘natural floodplains’ (no man-made activity)
prevails, the current vast areas, being used for agricultural and urban settlements, would
not be available. In the Indus plains, even though intensive structural measures are in
place, flow width of major rivers extends up to 20 km at most of the locations. These
widths would have reached many times of the present widths if these were not curtailed
by the present dike system. The structural measures facilitate coping with the everincreasing demographic pressure on the floodplains to meet their residential and food
needs by controlling floods (Ref. 3.2.1). So far, flood control has made developments
possible in floodplains with a focus on human health, safety and valuable goods and
property (White et al., 2001; APFM, 2007d). In addition to some environmental and
ecological issues associated with the structural measures, extended protection provided
by flood control attracts more investments in floodplains and increases the risk that was
reduced by the measures.
Controlling the vulnerability of a society is another way to reduce flood risk.
Unfortunately, vulnerability controlling methods are comparatively less developed and
sometimes more resource demanding as well as less effective. Measures that reduce
vulnerability, manage flood risk either by reducing susceptibility of elements at risk or
reducing exposure in floodplain. Reducing susceptibility is not always a cost efficient
solution. Whereas exposure control may reduce the substantial economic and societal
benefits of floodplain utilization (ASFPM, 2004). Whereas, the population pressure and
lack of other farmland emphasize on developing the floodplain (Plate, 2002). Therefore,
flood control might be considered as the first choice. Flood control can be performed in
the following ways:
Curtailing flood spread
Attenuating peak
Flow diversion etc.
Only those measures that help in curtailing flood spreads will be considered under the
scope of this PhD thesis.
Although, fixing rivers to a stable regime often fails and is usually expensive, it is
unavoidable in some circumstances (Green et al., 2000). However, the developments in
floodplains will continue (ASFPM, 2003) and ever-increasing protection will remain in
Design of structural measures
Present practices of design of structural measures are mostly probability-based (Di
Baldassarre et al., 2009). Generally, measures are uniformly designed nationwide to
withstand a design flood of specific return period.
Residual risk
Consideration of residual risk at the time of strategy formulation is very important. Floods
that result against residual risk are extremely catastrophic due to their intense nature and
unexpected occurrence. The residual flood risk behind levees or downstream of dams
remains mostly ignored and often is not considered into decision-making of land-use
planning. Risk-based assessments provide a better comprehension of these risks
(Duivendijk, 1999). There is a growing understanding that total security against floods is
impossible (Alkema and Middelkoop, 2007; APFM, 2007d). Absolute protection from
flooding is technically not feasible as well as economically and environmentally unviable
(APFM, 2009a). Whatever is done to prevent floods from happening, there will always be
some residual risk (Genovese, 2006). The residual risk is the portion of the risk that
remains even after flood measures have been initiated. Major sources of residual risk are
design assumptions and imperfect knowledge. For example, the probability of structural
failure against floods lower than the design flood (Carter, 2005), or even the hydrological
conditions in the catchment may change in time for various reasons (Duivendijk, 1999). A
zero level risk is not a realistic and consistent target as it would involve huge costs (Gilard
and Givone, 1997). However, it is important that floodplain inhabitants must be aware of
the residual risk (Ref. 5.1.1). Accordingly, floodplain inhabitants may decide to accept or
decline that risk against the benefits of the floodplain as a tradeoff.
In other words, deciding the design probability1 is a compromise, a compromise on safety
or a compromise on developments. The proposed ORP standards demonstrate the
existence of residual risk that may be termed as ‘floodplain land-use rent’. Part of this rent
is associated to costs of flood measures2 (if there is any) and rest is due to flood damages
1 A common mistake in practice is regarding the design of various types of utility-lines that may pass
through the floodplain. Their design probability is usually correlated with the river size. Whereas this
must be based on the functional importance of service line and involved costs, instead.
These costs are induced impacts of floods (Ref. 3.1).
(Figure 3-4 and Figure 3-5). Flood planning must consider the consequences of floods that
are more severe than the design flood, and must include the management of such floods
(APFM, 2009a), which can be performed by non-structural measures.
Induced risk
Structural measures alter the flow regime. Most of the measures just curtail flood spread
or divert peak flows. This suppression or diversion induces flood risk for areas on the
opposite bank, upstream or even sometimes to downstream areas. This shifting of risk to
other areas has often been neglected so far (Collins and Simpson, 2007; Majewski, 2007).
The second type of induced risk is due to increased intensity of floods when a structural
measure fails (dike, dams, etc.). In addition, the vulnerability of the society increases, as
people feel safer when flood protection is provided. High protection standards reduce
flood probability resulting in decreased public awareness and willingness to
cooperate/take responsibility for flood protection (Alphen and Lodder, 2006). The
induced flood risk must properly be modeled and cost-benefit analysis must cover the
resulting impacts in all areas involved (Ref. 3.6.1).
River dynamics
Rivers are dynamic systems whose forms vary according to variations in runoff and
sediment loads, over time (Green et al., 2000). As mentioned in subsection 5.1.2, the
structural measures may cause shifting of flood risk to upstream, downstream, or even
across the bank. In addition to shifting of risk, river flow pattern is also disturbed. While
designing a structural measure, the modified flow pattern must be considered to envisage
future behavior of floods. Commonly, hydrodynamic models are used with the help of
geographical data to visualize modified patterns of river. The spatial visualizations help in
designing flood measures (Carton, 2002). These visualizations are represented by plotting
various hypothetical flood parameters graphically using flood maps (Tariq, 2005). These
flood maps are used as the base for the assessment of consequences.
Socioeconomic development and flood management
Socioeconomic development and flood management in floodplains support the
development of each other. The theoretical level of risk that would prevail in the absence
of any flood measure is represented by the ‘baseline’ (Erdlenbruch et al., 2009). Most of
the floodplains carry economic activities and experience modified flow regimes after
having flood measures being incorporated. The purpose of defining a baseline state is to
evaluate the economic efficiency of the present state of flood management in a floodplain.
The baseline state is also needed as a reference to calculate the deviation of already
developed setup from the optimum scheme.
Some common experiences explain the public expectations and demands for the flood
Once floodplains become urbanized, these follow an almost inevitable demand
from the local community for flood protection.
Floods with higher intensities cause catastrophes are followed by public demand
for protection from floods (Tucci, 2002).
Interaction between embankment strengthening and socioeconomic developments is
explained by the ‘diking cycle’ (De Bruijn, 2005). The ‘diking cycle’ represents turn-byturn increments in safety levels and added exposure on repeated basis. When an area is
protected by constructing dikes, floodplain inhabitants feel more secure and raise their
investments in the floodplain (Di Baldassarre et al., 2009). This increase in exposure
demands higher safety, which is then provided by raising dike heights. Increments on
both sides continue turn by turn. Theoretically speaking, this cycles seems to have no end,
but very high dikes should be avoided (Tucci, 2007). The ideas of ‘Room for the River’ (in
the Netherlands), and ‘Making Space for Water’ (in the UK) are good practical examples to
overcome this problem. The references of baseline and ORP help in evaluating the best
options of achieving ORP at an existing state of flood management setup.
The flood resulting from a dike failure produces more severe impacts than if there were
no dike (De Bruijn, 2005; Tucci, 2007), possibly because of sudden and high flow and
secondly because the inhabitants are not prepared for the incident. Therefore, it can be
concluded that residual and induced risks cannot be avoided, but can be minimized and
must be considered while designing flood measures. These standards sometimes lose
their economic efficiency in an attempt to make structures withstand against design
floods. As a result, high safety factors and large freeboards result in uneconomical designs.
These make hydraulic structures more expensive to construct and maintain in good
conditions (Majewski, 2007). As a general principle, flood structures in floodplains should
be designed to fit within their socioeconomic bounds, and do not detract from the
harmony of the floodplain. In addition, flood protection standards must be consistent with
the available resources and the requirements for economic efficiency (Handmer, 1990;
Duivendijk, 1999).
The risk-based flood management and concept of ORP standards (Ref. 3.6.5) support a
proactive approach and facilitate decision-makers to decide the level of protection to be
provided. Conventional approaches do not guide in these situations and continue
increasing protection levels through ‘diking cycles’ (Ref. 5.1.4). This increased protection
encourages more exposure and reduced susceptibility that results in increased risk.
Risk-based design of dikes and dredging: Case studies
Risk-based design can be defined as the practice to evaluate the schemes for reducing but
not necessarily eliminating the overall risk (against affordable costs), as tradeoff between
the investment costs and the expected losses in case of failure (Tung, 2002; Pilon et al.,
2003). Risk-based design is performed according to the framework developed in Chapter
There are some segmented dikes in both study areas, which protect areas that are more
vulnerable. These individual dikes are constructed to prevent flooding at local level. These
dikes are built according to the prevailing design standards of 50-years return period
flood in Pakistan (Halcrow et al., 2001). In our case studies, the proposed design
standards (Ref. 3.6.5) are followed to optimize the design of the dike and dredging.
Detailed CBA of the baseline case and proposed designs are analyzed to achieve the
optimal state.
Dike-crest level design (the Swan River case)
For study purposes, two types of dikes are evaluated. In the first case, continuous dikes on
both sides of the river are introduced. In the second case, fragmented dike sections are
introduced to areas that are more vulnerable. Indication of vulnerable area in second case
is done based on a baseline EAD distribution map (Figure 5-3). The flood impacts against
existing dikes, the baseline case, continuous dikes, and fragmental dikes are evaluated as
well as compared against ORP. Dikes are introduced on the natural embankments, as per
standard practice (Duivendijk, 1999).
Continuous dikes of five different heights (ranging from 1m to 5m with one-meter
increments) were introduced1. Figure 5-1 represents the comparison of resulting damage
curves of baseline case, existing dike system, and proposed dikes. Damage curves of all
cases show complex trends due to the dual role of dikes. Dikes keep floodwater away from
the protected areas, unless the floodwater depths exceed the design heights of the dikes.
However, once floodwater overtops, dikes prevent floodwater to return to the river, hence
exacerbating the losses.
The terrain in the Swan River study area is steep. According to our results, dikes below
4m, did not play any significant role in flood prevention as flow stage rises rapidly up to
that level. Higher dikes (4m and 5m dikes) including the existing dikes (4m high)
efficiently control flood spreads against floods with 0.2-probability or lower. For further
severe (lower probability) floods, these dikes cause increase in flood losses by preventing
the return of flood spreads to the river. These dikes even lead this escaped floodwater
towards lowlands by driving it in the shape of parallel streams in protected areas. This
phenomenon, known as ‘shielding effect’, causes flood losses to be higher than the
baseline situation. The jumps in the damage curve are noticeable when floodwater depths
exceed the heights of 4m and 5m dikes (Figure 5-1).
To increase economic efficiency of the proposed dike system, fragmented dikes were
introduced to control wider spreads only. The locations of these dikes are decided by
inspecting the flood risk spread of the baseline EAD distribution map (Figure 5-3).
Fragmental dikes were proposed interactively with heights ranging from 1m to 5m. In the
upper reach of the study area, dikes were set farther apart to provide more space for the
river flow to facilitate safe passage of high floods. Resulting damage curves show dikes of
Detailed methodology of all involved steps has been described in Chapter 4.
less than 3m do not reduce flood losses (Figure 5-2). Loss reduction trends of these dikes
are not complex due to absence of shielding effect. Different heights of dikes control
flooding of different probabilities e.g., 3m for 0.2-probability flood, 4m for 0.02probability flood and 5m for 0.001-probability flood.
Figure 5-1: Damage curves of existing, baseline case, and continuous dikes of different heights show
the trends in flood losses for the Swan River.
Figure 5-2: Damage curves of existing, baseline case, and fragmental dikes of different heights show
the trends in flood losses for the Swan River.
Figure 5-3: EAD distribution maps of existing, baseline, fragmental dike, and
continuous dike (both at 4m heights) for the Swan River.
Results show that setting fragmental dikes farther apart in the upper reach to widen flow
path helps in controlling high floods (Figure 5-3). Flood losses are reduced significantly in
protected areas whereas flood losses in unprotected areas remain almost the same. As we
found in the continuous dikes case that dikes are not an economically feasible solution for
areas that are kept unprotected in this case study, flood zoning might prove helpful.
The case study produced some more interesting results. Having even the same heights of
dikes (4m), flood losses of continuous dikes were higher than the fragmental dikes due to
the so-called shielding effect (Figure 5-3). In addition, for very severe floods (probability
0.1 and lesser), the low height dikes (up to 3m high) play practically no role.
Results of our simulations may differ slightly as compared to the failure behavior of real
earthen dikes. Once floodwater overtops earthen dikes, a breach develops. This
phenomenon causes high flow rate of floodwater entering the protected area and
facilitates return of floodwater after the flood peak passes. In our simulations, only
surplus quantity of water overtops towards the protected areas. Although ‘Sobek’ can
facilitate breach simulations when breach locations are pre-assigned, this was neither the
case nor efficient due to the hundreds of simulations in our case studies. Hence, our case
studies may provide lower estimates of losses in case of earthen dikes simulations.
Therefore, instead of showing jumps, damage curves move upwards gradually after the
floodwater depth exceeds the height of dike.
The spatial distribution of EAD for all proposed dikes is shown in Annex A. The floodwater
mainly spreads at the upper reach where the terrain is relatively flat. In the lower reach,
flood spreads confined to comparatively limited areas. Such limited spreads are due to
steep geography of the study area. Flood losses in such areas may be avoided by
implementing flood zoning (Chapter 6) in an economically feasible way.
Dike-crest level design (the Chenab River case)
Continuous dikes are also introduced to the Chenab River study area. The floodplain is flat
and dikes may prove an effective solution. Differences in damage curves in Figure 5-4
show the gradual decrease in flood losses as dike height increases. With the increase in
dike height, the number of locations where floodwater overtops are reduced, which
results in loss reduction. More chances of flood spreads are just upstream of Alexandria
Bridge (Figure 5-5 and Figure 5-6) where the very important Grand Trunk road connects
Islamabad and Lahore via Gujrat and Wazirabad cities. One reason is that upstream
terrain level at the right bank of the river is very low. Secondly, dikes are narrowing at the
bridge location and available space for the flow is small. Two important cities (namely,
Gujrat at right and Wazirabad at left) are situated at this location. To reduce this flooding,
dikes should be set a bit further from the river.
Figure 5-7 demonstrates the reduction in EAD distribution over the floodplain with 5 m
high dikes. EAD distribution maps for all proposed dikes in the Chenab River study area
are provided in Annex B. As the dike height increases, flood losses decrease. Locations
with high losses may be treated with flood zoning and dikes may be made wider at these
locations. Another highly vulnerable location is the Sialkot International Airport.
Localized ring dikes may be provided to reduce part of the losses. However, the damages
might be still high as indirect losses are high for such facilities.
Figure 5-4: Damage curves of existing, baseline case, and continuous dikes of different
heights show the trends in flood losses against floods for the Chenab River.
Figure 5-5: EAD distribution maps of existing case show the spatial distribution of
EAD over the Chenab River reach
Figure 5-6: EAD distribution maps of baseline case show the spatial distribution of
EAD over the Chenab River reach
Figure 5-7: EAD distribution map of proposed dike showing the spatial distribution of
EAD against 5m high proposed dikes over the Chenab River reach
Figure 5-8: Damages curves of existing, baseline case, and continuous dredging of
different depths show the trend in flood losses for the Swan River
Figure 5-9: Damages curves of existing, baseline case, and continuous dredging of
different depths show the trend in flood losses for the Chenab River
Risk-based river dredging
Increasing the hydraulic conductivity of channel increases its flow carrying capacity.
Therefore, our second type of structural measure, which is applied in our study areas, is
dredging. Dredging takes place over widths of 100m (for the Swan River) and 1.5km (for
the Chenab River) while depths ranging from 1m to 5m are simulated. Figure 5-8 and
Figure 5-9 show the impact of dredging on flood losses in both areas. In both areas, losses
are reduced with the increase in depths. Figure 5-10 and Figure 5-11 show the EAD
distribution maps of the Swan River area and the Chenab River study area at 5m deep
dredging, respectively. Detailed EAD distribution maps for all cases can be found in Annex
C and D.
Figure 5-10: EAD distribution map of 5m deep continuous dredging over the Swan
River reach.
Figure 5-11: EAD distribution map of proposed 5m deep dredging over the Chenab
River reach.
Although both dikes and dredging help in reducing flood damage, dikes proved more
economical as compared to dredging as more efforts are involved in dredging. The high
dredging costs tremendously increase the total losses (Figure 5-12 and Figure 5-13,
crosses graph scale limits). Based on the costs involved in dredging, it is not a
recommended solution, but dredging can be achieved economically if the riverbed is used
as borrow pit for nearby construction projects.
Neither a continuous dikes nor dredging are feasible for the Swan River. Figure 5-12
demonstrates that a continuous dike of any height is not a feasible solution and cannot
achieve ORP. Therefore, it was decided to introduce dikes to only those areas where
floods occur often and cause more damages. Providing dikes with more space for high
flows results in reduced losses (Figure 5-12). Overtopping of floodwaters across the dikes,
mostly accompanying by growing breaches, can exacerbate damage. As a generalized
conclusion, increase in dike height reduces overall flood losses but dikes also increase
flood losses by blocking the return of floodwaters back to river. Figure 5-12 shows that
4m high fragmental dikes are the most economical efficient solution.
In the Chenab River case study, continuous dikes prove economically efficient in reducing
flood losses. High costs of dredging make it unfeasible for the Chenab River. Figure 5-13
shows that 5 m high dikes are the most economical solution. Further accuracy should be
achieved by assigning the probability of failure to dikes. Soil or sand excavation from the
riverbed for construction and earthworks may prove a cheap but an unorganized
dredging type that may be supplemented with organized dredging.
Figure 5-12: Relationships between the increasing efforts to curtail flood losses and
losses due to floods and costs spent on measures to determine ORP for the Swan River.
Figure 5-13: Relationships between the increasing efforts to curtail flood losses and
losses due to floods and costs spent on measures to determine ORP for Chenab River.
6 Vulnerability adjustments
Existing flood management strategies in most countries are primarily riveted to the
structural measures (dikes), enhancing only the threshold capacity of the community. The
importance of non-structural measures was recognized only much later as compared to
structural measures. Flood measures between 1850 and 1950 were largely comprised of
structural solutions in the USA (Myers, 1995), which changed flow regimes, loss of
habitats, biological diversity and productivity. It has been observed worldwide that both
flood damage and spending on structural measures are increasing. The structural
measures normally cause substantial damages to floodplain ecology and social setup of a
floodplain area. Concerns over increasing flood losses and floodplain ecological
degradations have been raised by flood managers all over the world and alternatives to
structural measures are getting more attention.
In addition to the environmental issues associated with the structural measures, the
advantages of non-structural measures should also be considered for flood management
strategies. Non-structural measures generally exhibit high economic efficiencies. Their
benefit-cost ratio may even be greater for developing countries as compared to structural
measures (Handmer, 1990). In addition, flood management by controlling the societal
vulnerability (non-structural measures) proves sometimes a potential option.
Another drawback of ignoring non-structural measures is that a unidirectional approach
may exhibit limited loss reduction capabilities. Appropriately interlinking of nonstructural and structural measures may increase the net benefits (Tariq et al., 2010c). For
a ‘diversity increasing vulnerability decreasing’ strategy, it is necessary to include all
components of vulnerability (de Graaf et al., 2007). Coupling non-structural measures
with structural measures, increases efficiency and often has proven cost-effective. Flood
management planning involves research into the ideal combination of these measures
(Tucci, 2007). Initial proposals for the use of non-structural measures to reduce flood
damages were made by the mid-1950's (Galloway, 2004). The trends of recently
developed policies seem thereby moving away from flood protection towards floodplain
adjustments (De Bruijn, 2005).Therefore, the impacts of non-structural measures
independently and in combination with structural measures are evaluated in this chapter.
The following measures for vulnerability management are evaluated:
Flood zoning
Early warning system
Rescue and relief etc
Only flood mapping/ flood zoning is evaluated for the optimization purposes.
Flood mapping
Structural measures support the growth of economic activities in floodplains and this
increased exposure then demands higher levels of flood protection. The high levels of
safety thus achieved remain not cost efficient due to the lower ratio of marginal benefits
to marginal costs at high safety levels. In addition, these also develop a false sense of
security by ignoring the raised residual risk at the same time. Overlooking the residual
flood risk (Ref. 5.1.1) is one of the main reasons for growing flood damages. People
erroneously or perhaps deliberately conclude that floods were ‘freak’ events of a premodern era, that will not reoccur now on, at least they will not experience in their lives
(Pottiera et al., 2005).
Reality is that floods can still occur and can wipe out the economic developments within a
floodplain (Pelling et al., 2004). Because total flood control is not possible (Ref. 5.1.1),
adjustments in social and economic exposure can be helpful to minimize the flood
damages. One solution is to minimize activities in floodplains. However, we must bear in
mind that flood-risk areas are also often attractive (APFM, 2007b) and sources of
livelihood. A complete abandonment of floodplains will deprive us these benefits. The
solution might be to restrain and adapt activities in the floodplains. Flood mapping and
flood zoning help to achieve these adjustments.
Flood maps are good tools to illustrate flooding behavior in the floodplains. A flood map is
a topographic map where the hypothetical flood characteristics are represented
graphically (Tariq, 2005). The purpose of flood maps is to show the areas that are subject
to flooding, the expected flood intensities, and, eventually, the different level of risks
within the flooded area (ASFPM, 2003). These maps provide guiding information to landuse planners, flood managers, and political decision makers. Governments may plan
possible reconstruction and rehabilitation programs based on these maps. These maps
may also be used for identifying critical projects that should receive priority in post
disaster assistance (Benson and Clay, 2004).
Types of flood maps
Flood maps can be used for a number of purposes. Consequently, their construction
techniques and types of parameters that they target to draw are also different. Flood
hazard maps contain information about the probability and magnitude of floods whereas
flood risk maps additionally provide the information about the consequences (Moel et al.,
2009). Flood risk maps are a valuable means of communicating risk to local communities
for their awareness, adaptation, and local empowerment (Alphen and Lodder, 2006). In
addition, flood hazard maps are the central instrument to facilitate interagency
cooperation for the planning and operational activities (APFM, 2007c). A classification of
flood maps is presented here.
Based on development technique
Historic flood maps are developed based on data collected after a flood has occurred.
These data are usually collected by local administration using public interviews or with
the help of signs of flooding, which can be seen on walls and trees.
Simulated flood maps are also commonly used. Numbers of hydraulic models are available
to compute flow in 1D and 2D (Ref. 4.1.3). These hydraulic models can produce quite
satisfactory results. These maps may help in probability based flood management
standards designs (Ref. 3.2.3).
Characteristics based classification
In general, maps may be used to express flood danger, risk, hazard, and vulnerability
Flood danger maps also known as inundation maps provide the least information. Mostly
these are flood extent maps against a predefined frequency or a historical flood event.
Sometimes these maps represent extents of multiple floods. These can be supplemented
with point information on other flood parameters (e.g. depth or velocity at some points)
and important exposed assets (e.g. hospitals, power stations).
Flood hazard maps provide flood hazard information in terms of the flood probability and
flood intensity being described by one or more parameters (Meon et al., 2006). On such
maps, the critical characteristics of a flood hazard, such as the depth and velocity of
flooding, should also be indicated (Green et al., 2000). These maps can also be used to
represent the flooding scenarios. As in the case of the Netherlands, flood maps are
developed on scenarios considering possible breaching at different sections. Flood hazard
maps must be produced for the areas where flooding could cause considerable damage
(Meon et al., 2006). Flood extents, depths, velocities, durations, propagations, and the
rates of water rising are important parameters and may be used depending on the
situation and the purpose of the map.
Flood vulnerability maps draw the vulnerability of inhabitants or developments in a
floodplain. Flood risk management targets information on the consequences of a flood.
The consequences of a flood depend broadly on the potential damage and the coping
capability (vulnerability) of a region to handle floods. Vulnerability distribution is difficult
to represent but a number of indicators can be used. Mostly, health, financial situation,
age, preparedness, transport facilities, ecological and social values, as well as early
warning mechanisms constitute the vulnerabilities.
Flood risk maps combine the hazard and vulnerability information at the same place. Risk
maps mostly cover direct damages that are calculated using intensity-damage functions.
Intangible damages can also be expressed in a qualitative way. An improved method has
been introduced in our research to express risk spread considering direct damages (Ref.
In modern flood management, the use of non-structural measures is popular and the use
of ‘flood risk maps’ is attracting more attention. A brief overview of current practices for
the comparison and analysis is provided as well as improvements are proposed in
subsections 6.3 and 6.4. The conventional approach of flood risk maps mostly generates
merely damage maps. These are drawn by combining the hazard maps and vulnerability
maps. Such maps fail to incorporate the probability considerations that are essential to
convert a damage map into a risk map. Flood risk maps must cover the impacts of all
possible floods and not just of one single design flood.
Flood mapping practices
In addition to the variations in development techniques, the regulations and
responsibilities associated to flood maps are also different all over the world. Details of
procedures and mapping techniques also vary from region to region due to local expertise
and concerns. Flood maps have been developed in many countries but with different
characteristics. All European Union member states are developing their flood maps
following the European Directives (European Commission, 2007). The directives require
the development of both of flood hazard and flood risk maps. Many of the EU countries
already had developed flood maps but now they are revising them under these directives
to meet the standards. Italy, Spain, and Switzerland already have official risk zone maps
(RWS and RIKZ, 2007).
Flood mapping in the USA
The Federal Emergency Management Agency (FEMA) is responsible of developing flood
maps for the USA. FEMA produces two types of maps, one for flood insurance and another
for emergency programs (FEMA, 2008). Flood maps are used for insurance, land-use
planning, emergency responses, public awareness and other purposes. Based on these
flood maps, evacuation plans have also been prepared (RWS and RIKZ, 2007).
The US Congress created the National Flood Insurance Program (NFIP) in 1969 to
regulate land-use management in floodplains by adopting and enforcing floodplain
management ordinances to reduce flood damages. They have introduced financial
mechanisms which encourage local societies to introduce construction limitations in highrisk areas (APFM, 2007b). NFIP targets controlling development in the floodway and at
elevations below the 100-years flood level (Galloway, 2004). The entire 0.01-probability
flood hazard area is considered as high-risk floodplain (ASFPM, 2003). Floodplain maps
are revised periodically in the USA (Collins and Simpson, 2007).
Flood mapping in the UK
In England & Wales, the Environment Agency has developed several mapping products to
raise awareness of flood risk and support decision-making. Flood maps are prepared for
both of river and sea floods. Flooding from rivers or the sea has been mapped for 1% and
0.5% probability floods respectively considering no defenses since defenses can be
‘overtopped’ or ‘fail’ if an extreme event flood occurs (EA, 2006). The locations of flood
defenses have been marked and the areas benefiting from defenses against the 0.01 and
0.005-probability floods are indicated. Areas being flooded by an extreme flood (0.01%
probability) are also shown on the map. The information on the actual (residual)
probability of flooding is presented in three categories used by the insurance industry in
the UK. Categories are not visible on map but can be perceived by clicking a location on
the flood map. Flood intensity and social vulnerability maps are also prepared but these
are not available on the internet (RWS and RIKZ, 2007). Flood intensities have been
calculated using depth, velocity, and debris concentrations.
Flood mapping in the Netherlands
In the Netherlands, the long-term project FLORIS (flood risk and safety in the
Netherlands) aims at estimating and mapping the probabilities and consequences of
flooding for all 53 dike rings in the Netherlands (Merz et al., 2007). Most of natural and
man-made risks are available officially on one website1 for the public and professionals.
Province-based single or combined disasters can be viewed on the website. The flood
maps show flood prone areas, as defined by more than one meter flooding depth with a
frequency larger than 1/4000 per year (RWS and RIKZ, 2007). In land-use planning, a
water opportunity map (WOM) is currently being used to outline the relationship
between water and land-use. Evacuation plans in relation to flood maps, have also been
prepared. Efficient, smooth, and fast evacuation is ensured by blocking the crossings and
enabling one-way traffic on evacuation routes.
Flood maps do not carry any associated legislation regarding the land-use planning.
Nevertheless, water opportunity map (WOM) and water assessment test (WAT) are
currently being used to outline the relationship between water and land-use and
consequences of any newly planned land-use development on the water (Voogd, 2006).
Flood mapping in Germany
The German Parliament adopted in July 2004 the ‘Flood Control Act’ after the severe
flooding in August 2002, (Merz et al., 2007). Federal states (Bundeslander) are
responsible for the generation of flood maps along with general flood management (Kron,
2007). Within the country, there are no standardized and uniform nomenclature or
agreed practices for the flood mapping (Merz et al., 2007). The flood maps are constructed
to represent different return periods (Moel et al., 2009). Flood hazard map guidelines of
the German Working Group of the Federal States on Water Issues (LAWA) recommends
the plotting of up to 1% probability flood or even less frequent (Meon et al., 2006). Public
flood-hazard maps are also available on internet for some states. Some states have also
provided the evacuation plans in event of flooding. Some states also show the hazard in a
qualitative way using a hazard matrix using intensity and probability of floods. New
developments are prohibited in areas with high flood risk outside existing settlements
(Wilke, 2004).
Flood mapping in France
In France, all natural disasters risks at any location can be found on a single interactive
map website so called cartorisque1. Flood maps are based on historical floods,
hydrogeomorphology, and flood modeling that is calculated against a single return period.
Qualitative risk maps have been created, which are usually classified into three to five risk
zones (Moel et al., 2009). Population, urban settlement, and infrastructure are used as
indicators for the exposure. There are some other websites, administered by different
agencies, showing flood maps developed by different methods (RWS and RIKZ, 2007).
Flood risk maps are developed for the insurance purposes, mainly. Legislation exists with
respect to restricting or prohibiting developments in flood-prone areas (Moel et al., 2009).
Development of new areas and modifications in areas that have already been urbanized
are separately treated (APFM, 2007b).
Flood mapping in Pakistan
The Federal Flood Commission of Pakistan started to develop flood mapping under FPSP
II in 1998 (Ref. 2.6.2). Initially, flood maps were developed for 5-years and 50-years
return period floods for major rivers of Indus basin. Initial maps have been developed,
whereas their calibration and validation is still to be done in the near future. The
development of maps has not been completed and their planned uses are not yet defined.
Development of official flood maps in most of developing countries either is not initiated
or otherwise still in initial stages.
Flood zoning
Flood zoning is the practical step towards implementation of response against floods by
adjusting vulnerabilities using the flood maps. Flood zoning consists of a set of rules for
the settlements on the floodplain with the aim of minimizing future damages due to floods
(Tucci, 2007). Future developments in floodplains must be regulated in order to control
risk. This can be achieved by establishing appropriate mechanisms that include legal
provisions defining the responsibilities for the damage control and social impacts
(Genovese, 2006). Flood zoning is the mechanism that regulates developments by dividing
a floodplain into zones and setting development regulations accordingly (ASFPM, 2003).
One purpose of zoning is to reduce the susceptibility of communities. Another purpose of
the zoning policies is to limit the number of vulnerable persons and assets in risky areas.
Governments are particularly interested to control possible expenditure associated with
the damage compensation (Erdlenbruch et al., 2009). Flood maps play an essential role for
all types of flood mitigation strategies. For example, building laws in Germany demand
that the flood risk in flood areas should not rise due to building activities (Wilke, 2004).
Unfortunately, there exists no uniform approach for the development and application of
flood risk maps (Smith, 1997; Pottiera et al., 2005; Merz et al., 2007). Flood risk maps are
being developed to estimate and alleviate the potential damages. Commonly, these maps
are based on hazard maps that show different zones corresponding to flood intensity in
different areas against a design flood. Such zones are subjected to different legislative and
guiding instruments that restrict or prohibit developments in certain zones. Sometimes
these regions are delineated based on floods of different return periods i.e., 25, 50 and
100-years (more common) (Apel et al., 2009) return period floods. This kind of zoning is
clearer than other types.
Spatial attributes of flood maps mainly depends on the geographic and hydrological
factors whereas ‘zoning regulations’ must be subject to the economic and social
conditions at the national level. Development and implementation of building codes for
the construction in floodplain can be an effective tool. Some countries have already
reformed their building codes through the flood management legislation (APFM, 2007c)
e.g., USA and Australia (NFRAG, 2008).
While establishing zoning regulations, it is important to consider the factors that force or
attract the society to live under flood risk. Particularly in developing countries, there are
often limited alternatives for economic growth and progress and the livelihoods of
millions of people depend directly on these natural resources. For example, in Pakistan, it
is the feudalism system that has forced the poor and powerless people into the floodplains
(White et al., 2001). In developing countries, population growth and migration towards
urban areas, mostly to the unplanned settlements in the floodplains, increase the
vulnerability of the poorest sectors of society to flooding (APFM, 2009a). The most
common situation, which encourages settlements in floodplain, occurs when a succession
of years with low floods takes place.
Resistance and reluctance against flood zoning
Implementation of structural measures is desired by the floodplain inhabitants. Nonstructural measures are generally not welcomed. As compared to other non-structural
measures, even more opposition is expected in case of land-use restrictions (Forster,
2008). Serious opposition and issues are expected in developing countries due to lack of
understanding about flood management, non-availability of alternate suitable land, and
expected relief and compensation commonly provided by governments in case of flood
losses (Kunreuther, 1996). Therefore, flood zoning must be introduced in situations
where it reduces losses and may efficiently raise land-use benefits.
Techniques associated with flood zoning tend to be less costly in capital expenditures but
more so in terms of human commitment (social capital) (Duivendijk, 1999). Rigorous
zoning regulations are usually violated at different levels of governments, societies, and
individuals (Moel et al., 2009), especially in developing countries. Merely restricting the
floodplain development to minimize the flood losses is sub-optimal. Regulation for the
urban settlements in floodplain areas must be facilitating and if possible, must be
discussed by the community before being implemented (Tucci, 2007). The objective of
flood zoning should be to enhance the floodplain land-use benefits. There is a certain need
to develop these flood risk maps in connection with the prevailing socioeconomic
conditions of society and their implementation should assure the overall advantages in a
rational way.
Risk-based flood zoning
Flood zoning based on a specified ‘design flood’ cannot be justified, as every floodplain
possesses unique hydrological, hydraulic, geophysical, environmental, and socioeconomic features. Doing so also violates the principle that “all floods must be managed
and not just some” (Green et al., 2000). In every floodplain, these properties along with
discharge regime must therefore be carefully considered while defining the flood risk
Inappropriately developed flood maps and inconsistent zoning regulations could be more
catastrophic and result in an inefficient land-use. Inadequate flood zoning, can restrain the
development, and may produce an illusion of intact safety to people living just outside the
delineated zone. This sense of security encourages increased exposure with high
susceptibility. A flood with intensity higher than the design flood may result in more
losses. Therefore, it is important to avoid hard delineation of zones and regulations must
be able to provide guidance to common land-use practices. Risk-based flood zoning counts
the actual benefits and drawbacks for any land-use.
At present, countries adopt different approaches towards flood zoning. A uniform and
universally accepted approach, which is still lacking, would allow comparison between
areas and facilitate communication between the regions by providing common bases
(Merz et al., 2007).
Flood zoning should be adjusted in a way that society may successfully reduce their losses
and remain be aware about the residual risk. As a rule of thumb, flood zoning should
consider to maximize the positive aspects of floods (APFM, 2009a). Risk-based flood
zoning not only provide a uniform basis but also facilitate to define regulations that suits
the socioeconomic conditions of floodplain inhabitants. Therefore, our main objective is to
increase net benefits in a floodplain but, strictly speaking, not to reduce flood losses.
Case studies
The risk-based approach was used to develop flood zones in both study areas. Risk
calculations were based on risk parameters defined in chapter 3. To cover the impacts of
all possible floods, ten scenarios were modeled with their probabilities ranging from 0.5
to 0.0001 (Ref. 4.1.1, 4.1.3, and 4.2). Damage functions developed by different sources
have been used in our damage model (Ref. 4.2.2 and Figure 4-7). Land-use maps (Ref.
4.2.1, Figure 4-5, and Figure 4-6) were used to represent the exposure. The risk equation
(Eq. 3-6) was used to calculate risk. Preliminary flood zoning for the dwelling (the Swan
River) and agricultural land (the Chenab River) were determined. Impacts of flood zoning
on the baseline case, existing, and proposed measures cases to achieve ORP, have been
As mentioned earlier, a risk-based approach naturally demands the consideration of
impacts due to all possible floods and not just one flood of a specific probability. Figure
6-1 represents annually expected virtual depths of all possible floods whose depths are
weighted according to their incremental probabilities of occurrences. Flood losses were
calculated against different land-uses. Expected annual damage (EAD) maps (Figure 6-3,
Figure 6-4, Figure 6-5, and Figure 6-13) were developed using Eq. 4-4. These EAD
distribution maps portray the spatial distribution of risk for different land-uses over the
floodplain. About ten land-use classes and six types of service lines were used in our case
Flood zoning for agricultural land
A change in land-use may have impacts on involved risk due to change in exposure (value)
and/ or susceptibility (damage function). Depending upon the damage function and
exposure of different land-uses, the most suitable land-use can be recommended for any
floodplain area. Owing to their significant role in flood damages, agricultural land in the
Chenab River area was selected for zoning purposes. Residual values of crop (rice) or
alternatively fodder (Figure 6-2) are calculated in the study area for zoning purposes. The
objective of zoning is set to increase the economic rent (especially location benefits ‘Rloc’
(Ref. 3.6.3)) in terms of net profit instead of just loss reduction. Location benefits are
obtained by deducting all involved costs of land-use and proportional costs involved on
flood management measures from the expected profit.
Figure 6-2 shows the residual net-profit for crop (rice) and fodder against inundation
depth. The net-profit for crop is high if there is no flood. Otherwise, impairing losses are
expected due to loss of investments involved for crop. On the other hand, pasture and
fodder are less profitable in no flood case, yet less impairing in case of flooding. The
reason is that less investment is involved for fodder and it is continuously harvested
throughout the season. A fraction of land is already harvested before it is flooded, hence
shows less flood losses.
Figure 6-1: A virtual flooding scenario showing weighted flood spreads based on their probabilities
of occurrence in study area, from Marala to Qadarabad, the Chenab River.
Figure 6-2: Relationship between flood depth and residual net-profit for agricultural land.
Figure 6-3: Maps show the distribution of expected annual net-profit of agricultural
land for the Chenab River reach against the existing conditions, before and after
Figure 6-4: Maps show the distribution of expected annual net-profit of agricultural
land for the Chenab River reach against the baseline, before and after zoning.
Figure 6-5: Maps show the distribution of expected annual net-profit of agricultural
land for the Chenab River reach against 6m dikes, before and after zoning.
Figure 6-6: Loss reduction economic effect of risk-based zoning for the Chenab River
Figure 6-7: Impacts of agricultural land-use zoning in reducing expected annual
damages for the Chenab River area.
Figure 6-3, Figure 6-4, and Figure 6-5 show residual net-profit of crop and low-profit/ loss
fields being replaced by fodder on locations where fodder generates comparatively higher
net-profit. Detailed results are presented in Annex E. Based on net profit, land-use
changes are proposed for all scenarios developed in subsection 5.2. The impacts on each
proposed flood management measure under defined flood zoning have been shown in
Figure 6-6.
Figure 6-8: Area selected in the 5km buffer of river centerline to evaluate the increase in net-profit
for the Chenab River case.
Figure 6-9: Increase in agricultural land-use net-profit for 5km buffer area around the Chenab
Main reduction in flood losses appears at 0.5-probability flood and afterwards this
reduction increases slightly for higher floods. It means that the area recommended for
land-use change is inundated almost every year. Figure 6-7 shows that the proposed flood
zoning contributes in reducing overall flood losses. The lowest point on the curve can be
used to achieve ORP after incorporating the impacts of early warning, storage, zoning for
more land-use options, and possible flow diversion options etc.
As discussed, the purpose of flood management in our research is to maximize location
benefits. The area delineated for flood zoning is in the proximity of the river. Therefore,
the impacts of zoning are concentrated in that area. An area within 5km buffer of the river
centerline is selected to demonstrate the impacts on net benefits (Figure 6-8). Figure 6-9
depicts the trends of net-benefits for all scenarios. Impacts of flood zoning decrease as
dike heights increase. Net-profit is maximum at 5m high dike.
Flood zoning for the settlements
Another important type of land-use is settlements. The Swan River area is selected, as the
area is situated in the capital city of Pakistan and is under high pressure for new
settlements. The study area is highly valuable and developers are interested to utilize
every inch of land. In such situations, proper flood risk evaluation becomes extremely
critical. Inappropriate land-use development may waste this land at one end or may lead
to high risk to life and dwellings, on the other hand. Utilizing the floodplains for
residential purposes in our case seems highly attractive, as the monthly-expected rent
from a house is high and the fact that floods do not destroy dwellings and possessions,
completely (Figure 4-7). Areas at a distance from the river are safe, as the land is not flat
like it was in the Chenab River case study. The areas that are in close proximity of the
river are under high risk and need proper evaluation of flood risk.
In our case studies, zoning was proposed to areas where EAD exceeded expected rent of
the residential area (Figure 6-13). Detailed results are presented in Annex F. These areas
may be used for any other suitable purposes. Alternate use of this land can be less
intensive land-uses e.g., park, car parking, picnic spot, or simply open space (as in our
case) that may store excessive floodwater, in case of high floods. Figure 6-10 shows a
slight decrease in EAD.
Figure 6-10: Impacts of dwellings land-use zoning in reducing expected annual damages for the
Swan River area
Figure 6-11: Floodplain area selected within 300m of the river centerline for the Swan River
case study.
Figure 6-12: Increase in residential land-use net-profit for 300m buffer area around the Swan River.
Figure 6-13: EAD distribution maps of baseline case, existing, and 4m fragmental dikes proposed
zoning for the Swan River reach
To elaborate the increase in net-profit, a narrower strip (300m) of the floodplain is
selected (Figure 6-11). Figure 6-12 shows the enhancement in net-profit of land-use after
zoning is applied to the Swan River study area.
Concluding remarks
Flood zoning not only helps in reducing the flood losses but also increases the land-use
net-benefits. Every flood measure has typical benefits along with some negative
characteristics. A synergetic combination is needed in this case. The selection of measures
depends on the nature of the flood problem and the target parameters (susceptibility,
exposure, intensity, and probability). The results of both case studies support the idea that
a combination of flood management measures may prove most efficient, economical, and
viable solution.
Detailed inspection of results reveals further significant facts. Land-use practices are
already complying with the flood considerations in areas that are situated in close
proximity of river. These areas experience floods frequently and flood damages are quite
substantial. No settlements or high value crops are cultivated in these areas. Therefore,
such areas, in our case studies, are considered as part of the river (Figure 6-3, Figure 6-4,
and Figure 6-5). Nevertheless, proposed flood zones are also frequently flooded areas,
therefore, more precision and accuracy is required to simulate high probability floods.
Structural measures prevent flood losses against floods up to the design flood. In contrary,
flood losses are minimized but not eliminated when flood zoning is implemented.
Therefore, it must be understood that the purpose of zoning here in our case is to increase
land-use benefits rather to eliminate risk.
As discussed, the zoning criterion in our case is to delineate the areas where present or
proposed land-use practice produce net-profits lower than achievable with even lower
investments. This does not mean that the area outside the marked zone is safe from
floods. Areas outside the marked zone still may suffer substantial flood losses but
expected net-profit is still positive and the maximum possible with the available
investments. In the Swan River case study, flood zoning delineates areas where dwelling
is not profitable.
Flood zoning is sensitive to structural measures and river processes. It may need to be
updated as time passes. Structural measures change the flow regime and may greatly
influence the design and impacts of non-structural measures. This fact is also supported
by our results as the layout of the flood zone changed significantly against different dike
heights. In addition, estimates of flow and stage are not static but need to be periodically
revised and updated. Simultaneously, vulnerability is also not constant in time due to
changes in social setup and economic growth. Therefore, risk maps must be updated
periodically to incorporate significant changes in flow and vulnerability.
Although flood zoning is effective to enhance land-use benefits, consideration of
intangible losses can enhance the reliability of results. The risk maps may assist
politicians and decision-makers in decision-support process. Although economic losses
are not the only type of losses and zoning is not the only option; a more effective and
comprehensive optimization can be performed by considering more possible flood
mitigation measures and all possible types of economic, social, and environmental
impacts. Overall risk reduction must be evaluated by optimizing the combined effects of
all suitable measures to define ORP.
Implementation of flood zoning requires an appropriate institutional support to guide and
convince inhabitants to adapt zoning recommendations. Farmers, who follow the zoning
recommendations, endure lower profit during no-flood periods, while others earn higher
profits. The same is true for encroachments for residential and commercial purposes.
Such investments in floodplains resemble gambling (Pottiera et al., 2005). That is why,
appropriate institutional support is needed that may ensure proper implementation of
flood zoning. If the area recommended for the zoning is less frequently flooded, then it
becomes even harder to convince the land-users to observe the flood zoning guidelines.
For such areas, less restriction or alternative methods should be proposed for effective
flood management. Conventional approaches of enforcing flood-zoning requirements are
not very effective in developing countries. A more effective setup is essential to
implement non-structural measures in general and flood zoning in particular.
7 Vulnerability indirect-adjustments
Chapter 6 demonstrated the positive impacts of flood zoning. Unfortunately, flood zoning
is not generally accepted by floodplain inhabitants, especially in developing countries.
Nevertheless, important benefits associated with flood zoning invoke the demand for its
implementation. To ensure appropriate implementation of flood zoning, the resistance to
zoning should be overcome by addressing these causes in an effective way. Normally,
enforcement of flood zoning is endorsed with legal and institutional arrangements. In
addition to all these efforts, tools that are more effective are needed for the achievement
of desired results. Risk-based flood insurance has been proposed as a possible way of
effective flood zoning enforcement. A risk-based insurance policy not only ensures the
enhanced coping ability of floodplain inhabitants but also supports flood zoning in
achieving desired land-use practices. Current flood insurance practices are not truly riskbased. With some modifications, conventional flood insurance mechanism can be
effectively shaped to support flood zoning.
Flood insurance in practice
Flood insurance is generally separated from the regular insurance. Mostly, all
extraordinary natural disasters are excluded from traditional insurance agreements. As a
result, property owners need to purchase additional policies in order to insure against
natural disasters (Andjelkovic, 2001; Collins and Simpson, 2007). Different countries
develop their own insurance objectives, development methodologies, premium
calculations, and implementation mechanisms according to their own priorities, technical
capabilities, and supporting institutional setups (Duivendijk, 1999).
Correlation between insurance policy and floods remains a debatable issue. In insurance
industry, risk assessment models are based either on flood zones (mostly based on a
design flood) or on calculation of flood losses for a series of historic events (Mehlhorn et
al., 2005). In addition, insurance rates are based on loss experience pooled nationally and
divided into properties located within the 100-years floodplain or coastal high-hazard
areas (Burby, 2001). The insurance program is expected to be self-supporting (i.e.
premiums are set at an actuarially sound level) in an average loss year, as reflected by
past experience (Collins and Simpson, 2007).
The implementation of flood insurance is mostly supported by legal provisions, public
incentives, and enforcement practices. For example, in the USA, no mortgage lenders that
are federally insured or financed can lend money on a property in a floodplain zone unless
the property is covered by flood insurance (Collins and Simpson, 2007).
Most significant issues are concerned with the public acceptability of insurance. An
admissible insurance rate is essential for the acceptability of insurance policy. Records
show that only about one in four homeowners, who live in a floodplains, purchase federal
flood insurance in the USA (Collins and Simpson, 2007). According to NFIP records, every
fifth policyholder discontinues flood insurance coverage each year (Kunreuther, 1996).
Several studies have found that only about 20% of those required to carry insurance
actually do so (Burby, 2001). Unfortunately, in developing countries, insurance is not an
option for the majority of people (APFM, 2006b).
General hesitation towards purchasing a flood insurance policy is due to the low
probability of flood events and response behavior of individuals, societies, and
governments. The effectiveness of an insurance approach suffers due to the public's
perception that flood will not come again and their expectation for disaster relief. If
insurance is available, then it is inequitable if the uninsured (and the grossly
underinsured) receive compensation as part of relief payments (Duivendijk, 1999). To
establish insurance policy, the social and political hindrances have to be appropriately
Flood insurance is usually provided to a limited area of a floodplain that is flooded
regularly, whereas, most flood losses stem from less frequent flood events (Burby, 2001;
Collins and Simpson, 2007). Reluctance to buy insurance is sometimes caused by the fact
that investing in insurance appears higher than flood risk itself (Kunreuther, 1996).
Improper calculation of premium results either due to lack of understanding, skills, and
data availability, or due to oversimplification of the estimated premium. The insurance
sector in developing countries is still weak and not in a position to calculate a realistic
premium to build up a flood insurance portfolio (Andjelkovic, 2001). Premium rates are
made uniform to larger areas in developed countries. For example, in the USA, the entire
1% flood hazard area is considered as high-risk floodplain (ASFPM, 2003). Premium rates
in the UK are broken down uniformly by postcode regardless of their exposure to the
flood hazard (Arnell, 1990) and these are often high (Collins and Simpson, 2007). A
mismatched insurance rate may reduce the acceptability of insurance policy and may
even prove counter-productive in overall flood management.
Risk-based flood insurance
In contrast to current practices, insurance premium rates must be proportional to
prevailing risk to describe the severity of the flood problem at any location. Flat rates or
rates calculated based on a single design flood do not adequately project the actual risk.
Risk-based flood insurance has been proposed in our case study to represent existent risk.
The approach, proposed in our research, is also compared with the conventional
Continuous population growth, improved living standards, urbanization, and
industrialization in floodplain areas are likely to enhance societal vulnerability to floods.
In the present age of science and technology, flooding cannot be regarded as an
unforeseeable event (Kron, 2002). The probability of flooding and its impacts can be
estimated accurately and precisely. Even today, post-flood management problems can be
planned in advance (Andjelkovic, 2001). In order to manage these expected problems
associated with floods, establishment of an appropriate setup is essential. The importance
of flood insurance increases manifolds in these situations. The flood insurance is a
versatile and complementary tool to deal with floods in many respects. Flood insurance
differs sharply from the other measures available for managing flood losses. As commonly
practiced, measures reduce the cost of damages from each flood, insurance mechanism
distributes the losses over time (Duivendijk, 1999).
Recovery from flood
The conventional approach to flood insurance mainly addresses the issue of offering
compensation for the losses caused by floods when damages are not avoidable at
acceptable costs. The flood insurance mainly spreads the cost of flood damage both in
terms of time in order to facilitate the affected society to deal with the aftermath of flood
events (Arnell, 1990; Duivendijk, 1999; Andjelkovic, 2001; APFM, 2007a). A flood strike
may disrupt economic activities in the affected area. Sometimes, delay in the economic
recovery of the affected people can cause extended indirect losses and they become a
constant burden on society. Proper insurance can help considerably in mitigating the
effects of floods and prevent flooded societies from being ruined (Kron, 2002).
Flood risk awareness
Flood risk awareness has been proved an effective measure in reducing negative impacts
of floods. The holding of community-based flood awareness programs may help in raising
flood awareness among the people. Unfortunately, flood management is reactive and not
proactive worldwide. Before a community experiences a flood, flood risk awareness
programs may not be very effective and may not produce the required results.
Devastations caused by flood events raise people’s attention to the problem. Surprisingly,
people often forget their proneness in a short time after a flood they have actually
experienced (Kron, 2002). In developing countries, even the public awareness programs
are not conducted regularly. Consequently, land-use planning regulations are mostly not
followed strictly, despite much efforts undertaken (Kron, 2002). The floodplain
inhabitants must be aware of the risk they possess. Flood insurance reminds people on a
regular basis about the risk they have accepted. The flood insurance, with rates depending
on the degree of the threat, is increasingly used to offset the threat (Majewski, 2007).
Reduction in damages
Insurance premiums act as regular reminders of hazard and produce risk awareness.
Ultimately, this risk awareness can be used in risk reduction. Reduction in risk is only
possible when insurance premiums are in proportion to actual risk and flood damages are
shared by the insured persons as well. If the insurance premiums are based on actual risk
then policyholders will avoid activities that may cause increase in premium rates (or risk
in other words). To reduce premium rates and to increase net-benefits of land-use,
policyholders will devise and adopt those land-use practices that must produce high netbenefits. Within the floodplain, different land-uses may range from highly beneficial to
extremely damaging. Sometimes, a complete abandonment is required and the other time
a small modification in land-use can produce the desired results. Flood insurance can help
to achieve the following goals:
Discouraging a land-use
Modifying a land-use
Promote a certain land-use
Replace one land-use by another
Higher premiums could be an effective deterrent to uneconomic floodplain encroachment
(Arnell, 1990). In the USA, the insurance program aims at limiting development in the
floodway (Galloway, 2004).
Insurance may have both advantageous and disadvantageous effects upon flood loss
potential (Arnell, 1990). Negative aspect of insurance is that people do not put serious
effort into protecting their insured assets. This shortcoming can be overcome by a partial
compensation of flood losses as is a usual insurance practice.
Fine-tuning in insurance premiums calculation methodology may lead to a number of
advantages. As highlighted earlier in this section, if insurance premium rates are based on
intensity of flood risk, above mentioned objectives can be obtained. In other words, riskbased insurance may shape land-user’s activities to restrain flood losses and increase
economic rent of land.
Advantages of risk-based insurance
In insurance industry, calculation of insurance premium is one of the most critical tasks to
accumulate maximum possible number of clients and to avoid adverse selection. Insurers
need precise and accurate flood hazard information in order to define realistic premiums
for flood damage insurance (Meon et al., 2006). Not all the communities within a
floodplain experience the same level of risk. It would be unfair and inexplicable to clients
if each member of an insured community had to pay the same premium without taking
into account the individual risk his property is exposed to (Kron, 2002). Therefore, the
premiums of flood insurance must also be according to the level of exposure.
Risk-based flood insurance can help in overcoming these issues. There is a need to
determine risk and collect insurance premium based on actual risk involved. At present, in
the absence of risk-based insurance setup, premium is collected on the basis of a flat rate
conventional approach. People in high-risk areas are subsidized by people in low-risk
areas (Crichton, 2006). Risk-maps can be used for this purpose as these maps portray the
actual distribution of risk and facilitate mitigating flood hazard on rational basis.
Insurance premiums based on a single flood event, not only collect disproportional
premium rates but also fail to provide insurance to areas under low risk. An effective and
appropriate insurance rate should be proportional to involved risk. Non-representative
flat rate will cause overcharging in low-risk areas and vice versa. If another insurer offers
low insurance rates, clients would obviously prefer to go for low rates. As a result, the
insurer will get more clients from high-risked areas where insurance premium rates are
lower than the potential damages. The insurer will bear the difference and will make a
loss. This phenomenon is known as ‘adverse selection’, ‘negative selection’, or ‘antiselection’.
Case study
Following the definition of risk-maps in 6.2, these maps help in determining flood impacts
over time and space frames and provide guidance to land-use planners, flood managers,
and political decision makers (Tariq et al., 2010b, a). Therefore, flood maps are used in
our case study for risk distribution assessment. With the combination of flood simulation
and damage modeling, EAD distribution maps are developed to represent the spatial
distribution of risk for agricultural areas (Ref. 4.3.4 and Figure 7-1). For the purpose of
comparison with the conventional approach, another loss map is also developed against
design flood (USA customary design flood with 1% probability, see Figure 7-2). The
impacts of both approaches on insurance-providers and insurance-holders have been
evaluated. A design flood approach provides the consequences against a specific flood and
may ignore the possible losses due to other floods. Damages under design flood were
adjusted to the actual EAD calculated considering all possible floods.
Results and conclusions
Results show major differences in spatial distribution of risk when damages of the 100years design flood are compared with EAD. The obvious difference is that the design flood
inundates the floodplain partially (compare Figure 7-1 and Figure 7-2). If flood losses due
to all floods are pooled and divided only on this much area, these will estimate insurance
premium higher than the actual risk. In addition, due to partial inundation of floodplain by
design flood, insurance service cannot be calculated for all areas in a floodplain.
Figure 7-1: Spatial distribution of flood risk considering all probable floods.
Figure 7-2: Spatial distribution of flood losses due to 1% exceedance flood.
Figure 7-3: The calculated damage against 1% probability design flood on all
stations in study area.
Figure 7-4: The calculated risk calculated damages considering all probable floods
on all stations in study area.
Another significant observation is that the damage distribution intensity is almost
uniform (and maximum) throughout inundated area in case of design flood (Figure 7-3),
whereas, risk intensity varies predominately over the floodplain. Flood risk normally
decreases as one move away from the river (Figure 7-4). The reason for this difference is
the fact that not all floods inundate the entire floodplain.
To elaborate results further in detail, average damage distributions at ten selected
sections along the reach were analyzed. Graphs (Figure 7-5 and Figure 7-6) show that the
distribution pattern of damage due to the design flood is different from the actual risk
distribution. Average flood damages of design flood are higher than the risk on the left
riverbank and lower on the right riverbank. This means that a design flood cannot
represent the risk. Furthermore, generally dikes constructed on the left bank shift the risk
towards the right bank (Figure 7-7). As dikes are designed in Pakistan to protect a flood of
a return period of 50-years, the design flood, which is a 100-years return period in our
case study, overtops these dikes and does not account for this shift in flood risk. There are
a number of infrastructures, like highways and railways that play an important role in risk
distribution. In case of design flood, only the effects of those structures are noticeable that
curtail design flood.
Figure 7-5: Comparison of estimated average damages based on 1% probability
flood with calculated damages considering all probable floods.
Figure 7-6: Comparison of estimated cumulative damages based on 1% probability
flood with calculated damages considering all probable floods.
Figure 7-7: Schematic cross section of the Chenab River elaborating the general
slope of ground and shifting of risk due to dikes which are mostly provided on leftbank side of the River.
To promote flood insurance in developing countries, a state sponsored comprehensive
program must be initiated. Effective implementation of the program can be achieved
through public awareness, attractive incentives, legal support, and strict compliance.
Subsidized insurance rates with governmental support must be introduced along with
public information campaigns. Assistance provided in case of natural disaster must be
denied if flood insurance is not purchased and maintained.
Contrary to the existing practice, calculations of insurance premium should be based on
actual risk and must be closely linked to land-use management practice. Insurance rates
must be calculated by giving due consideration to the dynamic nature of river flows and
land-use practices. Risk reduction by land-use modifications must be appreciated in order
to promote zoning. This would also improve the risk pooling efficiency by addressing
simultaneously rural and urban sectors.
The basic financial flow conception of flood insurance is based on self-reliance ideology.
Normally, the issues relating to ‘who reaps the benefits and who pays’ often hinder the
generation of required resources for flood management (Green et al., 2000; Andjelkovic,
2001). The biggest hurdle on the way to set up effective flood management in developing
countries is their unnecessary reliance on others to provide assistance. Therefore,
whether on local or on national scale, flood management must be self-supported.
8 Conclusions and recommendations
Flood management affects society at a large scale. Society and its functioning are at stake;
hence, the socioeconomic aspects of the flood management cannot be ignored. As long as
the population, living standards, urbanization, industrialization, and high yielding
agriculture continue, the demands for enhanced and effective flood management will also
raise. Simultaneously, enormous care, appropriate resources, broad-based
multidisciplinary approach, and comprehensive understanding are required to handle the
issue. This thesis provides the basic understanding of risk-based flood management.
Although, focus of this research lie on developing countries, most results are equally
applicable to developed countries.
Discussion and conclusions
By supporting agricultural economies, fertile floodplains in developing countries are vital
sources of economic activities and cherished locations for settlements. Flood losses are
regular setbacks for their economies. Flood management in Pakistan is a typical example
of flood management in a developing country due to the diverse nature of the flooding
problem in the country. The study of Pakistan’s flood management system supports and
urges on the need of the understanding the flooding behavior and societal response
before developing a flood management strategy for a country.
Neither a conventional approach of flood control nor a simple replication of modern
methods of developed countries can solve the problem. A tailor-made flood management
approach is necessary to address local flood problems appropriately. Nevertheless,
experiences of developed countries can be employed for the improvement of flood
management in developing countries. Simple duplication of strategies, measures, plans,
and safety standards of developed countries is rather counterproductive. There cannot be
a single generalized solution applicable to every floodplain but there can be a single
acceptable approach for the implementation of consistent and uniform design standards.
Developing countries must develop flood standards according to their own socioeconomic conditions. To shape the most suitable flood management strategy, its impacts
and people’s response must be envisaged be forehand (Ref. 3.3). This is important to have
an idea about the expected outcomes of flood management, whether positive or negative,
as no strategy is perfect and able to eliminate all losses.
Flood events are part of nature. They have existed and will continue to exist. Economic
flood management will never preempt all future flooding but the damages by probable
future flooding must be brought into account in the design and cost of flood management
schemes. Such approach must conform to the local socio-economic context and define the
acceptable risk accordingly (Vrijling et al., 1998, 2000). The acceptability of risk is a
function of social, economic, and environmental concerns and may vary largely from
nation to nation. The commonly stated connection between acceptable risk and
probability-based design standards is rather tenuous (Ref. 3.2.3). The risk-based
standards are not only based on the acceptable risk but also have the capability to provide
the best possible solutions considering the available options and future risks.
In accordance with the proposed risk mechanics (Ref. 3.4), flood management can be
categorized into two distinct categories; namely prevention and relief or alternatively
hazard adjustments and vulnerability adjustments. Conventional approaches towards
flood management mainly comprise of measures that try to adjust only hazard. Our case
studies supported the fact that vulnerability adjustments may further reduce the risk. In
other words, the optimization of flood management cannot be accomplished without
tuning both ends.
It is extremely important to facilitate the decision-support process by providing correct
and sufficient information. Therefore, the standard methods should assure that results
portray the actual situation in the most convenient way. Flood mapping acts as a visual
language and provides support that is readily understood by policymakers. Mapping
should be an integral part of flood management evaluation. Moreover, up-to-date flood
information that is consistent over the entire territory is required for an effective
implementation of flood management strategy.
So far, the main unaddressed issue associated with the risk-based approach is its inability
to render uniform flood management standards. Such uniform standards provide fair
justifications for the flood management strategies. The probability-based standards have
an edge in this regard. Our research attempts to address this problem and the concept of
‘optimal risk point’ (ORP) has been proposed. The flood management standards at
national level can be established by correlating to the optimized state of a floodplain. Riskbased uniform standards may facilitate establishing economic, social, and environmental
Although, this research provides the guidelines for the practical implementation of a riskbased approach, this exercise itself is an academic activity. Further case studies are
required to enhance its practical utilization. Additional case studies consisting of
additional measures on more floodplains are required to generalize the findings of this
research. Implementation of this research to practical problems must be supported with
adequate social, environmental, and economic expertise. Uncertainty analyses are also
required while taking concrete steps.
Although, the research uses the risk-based principle, the ‘beneficiary pays the costs’1
principle is a basic constituent working in the background. Flood management can be
optimized to achieve minimum losses, maximum protection, maximum BC ratio, or
maximum land-use benefits. The optimization parameter in this research is ‘maximum
land-use benefits’ by reducing ‘flood deductions’ (Ref. 3.6). Therefore, resources required
for flood management can be generated from water users and floodplain inhabitants.
Although, redistribution of risk is inevitable to reduce overall risk, suffering stakeholders
must be compensated by those who benefit from such redistribution. Ignoring this
principle may severely reduce the acceptability and eventually the efficiency of flood
management strategy.
Flood management is an integral part of the land-use development. An improvised form of
flood management, called ‘floodplain management’ must be introduced. Floodplain
management should cover the river management and promote the coordinated
development and management of land-use resources. Just aiming at lowering the damages
and ignoring possible land-use benefits is rather counterproductive and must be avoided.
More reliable and consensual approaches must be developed to incorporate social and
environmental damages into risk calculations. Valuation of social and environmental
assets into monetary terms should not be carried out on individual case basis. Instead, a
comprehensive valuation ‘one-for-all-purposes’ at the national level must be performed to
ensure the implementation of fair and consistent standard for any project and any area in
a country. The governments need to establish supporting institutional setups for this
Flood management must be financially self-supporting at all levels of setups. In
developing countries, the prominent hurdle in the way of establishing comprehensive and
sustainable flood management setup is their reliance and dependency on others to
provide assistance at the eve of flooding and flood management projects. The principle of
‘beneficiary pays the costs’ can be adopted to efficiently generate funds for flood
management. Another potential problem in the aftermath of disasters in developing
countries is that in the rush to fast-track assistance, new projects are not properly
designed or appraised. A proactive approach is highly recommended. Evaluation of ORP
The word ‘pays’ is to be understood in welfare economic terms, it also covers the sufferings from
one’s actions. For example, floodplain inhabitants are affected when dikes are raised on the opposite
bank, or the general public suffers from the pollution caused by an industry at large scale or by a car
at small scale etc. In such situations, these groups must be compensated by those who are getting
benefits from these dikes, factory, or car.
for almost all floodplains must be carried out in advance. Emergency responses and
measures required afterwards should also be planned.
Freedom of land-use choice of floodplain inhabitants must be honored. Recommendation
about suitable land-use is a constructive step but restrictions and compulsions must be
avoided as far as possible. Personal choice of accepting risk may be covered by
introducing complementary risk-based insurance or flood tax. This can facilitate a
harmonious coexistence with floods.
ADB, 1997. Guidelines for the economic analysis of projects. Asian Development Bank.
ADB, 2008. Pakistan: Second flood protection sector project, in: Leonard, R.K. (Ed.).
Operations Evaluation Department, Asian Development Bank, p. 137.
ADPC, 2005. Integrated flood risk management in Asia, Bangkok, Thailand, p. 313.
Ahmad, I., 1994. Design Flood, PC based hydraulic design packages, 1.2 ed, p. A flood
hydrograph PC based package.
Alkema, D., Middelkoop, H., 2007. The Influence of floodplain Compartmentalization on
flood risk within the Rhine-Meuse Delta, in: Begum, S., Stive, M.J.F., Hall, J.W. (Eds.), Flood
Risk Management in Europe: Innovation in policy and practice. Springer, Dordrecht, the
Netherlands, pp. 21-42.
Alphen, J.v., Lodder, Q., 2006. Integrated flood management: experiences of 13 countries
with their implementation and day-to-day management. Irrigation and Drainage 55, 159171.
Andjelkovic, I., 2001. Guidelines on non-structural measures in urban flood management,
Technical Documents in Hydrology. UNESCO, International Hydrological Programme,
Paris, France.
ANFAS, 2003. Literature review for a socio-economic impacts assessment procedure for
Qianliang Hu Detention Basin, Hunan Province, China, Yangtze River Flood Control and
Management Project, p. 37.
Apel, H., Merz, B., Thieken, A.H., 2009. Influence of dike breaches on flood frequency
estimation. Computers & Geosciences 35, 907-923.
APFM, 2006a. Environmental aspects of integrated flood management, Flood Management
Policy Series. Associated Programme on Flood Management, Geneva, Switzerland.
APFM, 2006b. Legal and institutional aspects of integrated flood management, Flood
Management Policy Series. Associated Programme on Flood Management, Geneva,
Switzerland, p. 91.
APFM, 2007a. Economic aspects of integrated flood management, Flood Management
Policy Series. Associated Programme on Flood Management, Geneva, Switzerland.
APFM, 2007b. Guidance on flash flood management: Recent experiences from central and
eastern Europe.
APFM, 2007c. The role of land-use planning in flood management: A tool for integrated
flood management, Flood Management Policy Series. Associated Programme on Flood
Management, Geneva, Switzerland.
APFM, 2007d. Strategy for flood management for Kafue river basin, Zambia. Associated
Programme Flood Management, p. 155.
APFM, 2009a. Integrated flood management Concept paper. World Meteorological
Organization, Geneva, Switzerland.
APFM, 2009b. Risk sharing in flood management: A tool for integrated flood management,
Flood Management Policy Series. Associated Programme on Flood Management, Geneva,
Army Corps of Engineers, U.S., 2011. The Hydrologic Engineering Center, Hydrological
Model System (HEC-HMS), Institute for Water Resources. Department of US Army Corps
of Engineers, Davis.
Arnell, N.W., 1990. Flood insurance and floodplain management, in: Handmer, J.W. (Ed.),
Flood hazard management: British and international perspectives. Geo Books.
ASFPM, 2003. No adverse impact: A toolkit for common sense floodplain management, in:
Larson, L.A., Klitzke, M.J., Brown, D.A. (Eds.). Association of State Floodplain Managers, p.
ASFPM, 2004. Reducing flood losses: Is the 1% chance (100-year) flood standard
sufficient?, Gilbert F. White National Flood Policy Forum 2004 Assembly, Washington D.C.
ASTER, 2011. http://www.gdem.aster.ersdac.or.jp/index.jsp.
Awan, S.A., 2003. Integrated flood management case study Pakistan: Flood management River Chenab from Marala to Khanki. The Associated Programme on Flood Management,
p. 30.
Baig, M.A., 2008. Floods and flood plains in Pakistan, 20th International Congress on
Irrigation and Drainage, Lahore, Pakistan.
Benson, C., Clay, E.J., 2004. Understanding the economic and financial impacts of natural
disasters. The International Bank for Reconstruction and Development / The World Bank,
1818 H Street, NW, Washington, DC 20433, Washington.
Blong, R., 1996. Volcanic Hazards Risk Assessment, in: Scarpa, R., Tilling, R.I. (Eds.),
Monitoring and Mitigation of Volcano Hazards. Springer-Verlag Berlin Heidelberg New
York, New York, p. 23.
Borrows, P., Bruin, D.d., 2006. The management of riverine flood risk. Irrigation and
Drainage 55, S151-S157.
Bruijn, K.M., Green, C., Johnson, C., McFadden, L., 2007. Evolving Concepts in Flood Risk
Management: Searching for a Common Language, Flood Risk Management in Europe, pp.
Buchele, B., Kreibich, H., Kron, A., Thieken, A., Ihringer, J., Oberle, P., Merz, B., Nestmann, F.,
2006. Flood-risk mapping: contributions towards an enhanced assessment of extreme
events and associated risks. Natural Hazards and Earth System Sciences.
Burby, R.J., 2001. Flood insurance and floodplain management: the US experience. Global
Environmental Change Part B: Environmental Hazards 3, 111-122.
Carter, N.T., 2005. Flood risk management: Federal role in infrastructure, CRS Report for
Congress. Congressional Research Service, p. 12.
Carton, L., 2002. Strengths and weaknesses of spatial language: Mapping activities as
debating instrument in a spatial planning process, FIG XXII International Congress,
Washington, D.C. USA, p. 13.
Cheema, M.J.M., Bastiaanssen, W.G.M., 2010. Land use and land cover classification in the
irrigated Indus Basin using growth phenology information from satellite data to support
water management analysis. Agricultural Water Management 97, 1541-1552.
Chetty, K., Smithers, J., 2005. Continuous simulation modelling for design flood estimation
in South Africa: Preliminary investigations in the Thukela catchment. Physics and
Chemistry of the Earth, Parts A/B/C 30, 634-638.
Chow, V.T., Maidment, D.R., Mays, L.W., 1988. Applied Hydrology. McGraw-Hill, New York,
Collins, E., Simpson, L., 2007. The impact of climate change on insuring flood risk, Biennial
Convention, Institute of Actuaries of Australia, Christchurch, New Zealand.
Contijoch, M., 2008. Balochistan small scale irrigation project, in: Bernard, C.A., Brizzi, A.
(Eds.), Irrigation and drainage. World Bank, p. 137.
Crichton, D., 1999. The Risk Triangle, in: Ingleton, J. (Ed.), Natural Disaster Management.
Tudor Rose, London, p. 2.
Crichton, D., 2006. Hurricanes and hydro-meteorology, in: Centre, B.H.R. (Ed.), ASSAL XVII
Annual Meeting, Lisbon, UK.
Cruz-Reyna, S.D.L., 1996. Long-Term Probabilistic Analysis of Future Explosive Eruptions,
in: Scarpa, R., Tilling, R.I. (Eds.), Monitoring and Mitigation of Volcano Hazards. SpringerVerlag Berlin Heidelberg New York.
Cuny, F.C., 1991. Living with floods : Alternatives for riverine flood mitigation. Land Use
Policy 8, 331-342.
DCLG, 2006. Planning Policy Statement 25: Development and Flood Risk. Department for
Communities and Local Government, London, UK.
De Bruijn, d.K.M., 2005. Resilience and flood risk management: a systems approach
applied to lowland rivers, Civil Engineering Department. Technische Universiteit Delft,
Delft, p. 210.
de Graaf, R., van de Giesen, N., van de Ven, F., 2007. Alternative water management
options to reduce vulnerability for climate change in the Netherlands. Natural Hazards.
Di Baldassarre, G., Castellarin, A., Montanari, A., Brath, A., 2009. Probability-weighted
hazard maps for comparing different flood risk management strategies: a case study.
Natural Hazards 50, 479-496.
Dilley, M., Chen, R.S., Deichmann, U., Lerner-Lam, A.L., Arnold, M., Agwe, J., Buys, P.,
Kjekstad, O., Lyon, B., Yetman, G., 2005. Natural disaster hotspots: A global risk analysis,
Disaster Risk Management Series. The World Bank, Hazard Management Unit,
Washington, D.C. USA, p. 131.
Douben, K.-J., 2006. Characteristics of river floods and flooding: A global overview, 1985–
2003. Irrigation and Drainage 55, 9-21.
Duivendijk, J.v., 1999. Assessment of Flood Management Options, Assessment of Flood
Control and Management Options. World Commission on Dams.
Duivendijk, J.v., 2006. The systematic approach to flooding problems. Irrigation and
Drainage 55, S55-S74.
Dutta, D., Herath, S., Musiake, K., 2001. Direct flood damage modeling towards urban flood
risk management, Joint Workshop on Urban Safety Engineering 2001, Asian Institute of
Technology, Bangkok, Thailand.
EA, 2006. Understanding Flood Risk Using our Flood Map; Identifying and understanding
Flood Risk in England & Wales. Environment Agency, Bristol.
EEA, 2003. Mapping the impacts of recent natural disasters and technological accidents in
Europe. European Environment Agency, Copenhagen, Denmark.
Erdlenbruch, K., Thoyer, S., Grelot, F., Kast, R., Enjolras, G., 2009. Risk-sharing policies in
the context of the French Flood Prevention Action Programmes. Journal of Environmental
Management 91, 363-369.
European Commission, 2007. Directive 2007/60/EC of the European Parliament and of
the Council of 23 October 2007 on the assessment and management of flood risks, in:
Parliament, E. (Ed.), 2007/60/EC.
FEMA, 2008. Flood Insurance Manual, in: Agency, F.E.M. (Ed.).
FFC, 2007. Annual flood report 2006. Federal Flood Commission Pakistan, Islamabad,
Pakistan, p. 36.
FFC, 2009. Annual flood report 2008. Federal Flood Commission Pakistan, Islamabad,
Pakistan, p. 57.
Forster, S., 2008. An analysis of hydraulic, environmental and economic impacts of flood
polder management at the Elbe River, Faculty of Mathematics and Natural Sciences.
University of Potsdam, Germany, Potsdam.
Framji, K.K., Mahajan, I.K., 1969. Irrigation and drainage in the world; a global review, 2
ed. International Commission on Irrigation & Drainage, New Delhi, India.
Francés, F., Botero, B.A., 2007. Flood frequency analysis for extreme events, in: Ashley, R.,
Garvin, S., Pasche, E., Vassilopoulos, A., Zevenbergen, C. (Eds.), Advances in Urban Flood
Management. Taylor & Francis/Balkema, pp. 123-138.
Galloway, G., 2004. Integrated Flood Management, Case Study, USA: Flood Management Mississippi River, WMO/GWP Associated Programme on Flood Management.
Genovese, E., 2006. A methodological approach to land use-based flood damage
assessment in urban areas: Prague case study. Institute for Environment and
Sustainability, Ispra, Italy.
Gilard, O., Givone, P., 1997. Flood risk management: new concepts and methods for
objective negotiations, in: G.H., L. (Ed.), Destructive Water: Water Caused Natural
Disasters, their Abatement and control. IAHS Press, Anaheim, California, USA.
Granger, K., Jones, T., Leiba, M., Scott, G., 1999. Community risk in cairns: A multi-hazard
risk assessment, Cities Project, Department of Industry, Scienceand Resources, Australia.
AGSO (Australian Geological Survey Organisation).
Green, C.H., Parker, D.J., Tunstall, S.M., Berga, L., 2000. Assessment of flood control and
management options, WCD Thematic Reviews Options Assessment: IV.4. World
Commission on Dams, Cape Town, South Africa.
Haider, N., 2006. Living with disasters: Disaster profiling of districts of Pakistan. National
Disaster Management Authority Islamabad, Pakistan, p. 65.
Halcrow, 2004. River Dove strategy scoping report. Environment Agency, the UK, p. 63.
Halcrow, Euroconsult, NDC, 2001. Capacity building for integrated river management and
subprojects implementation design criteria and methodology: Package B Second Flood
Protection Sector Project II. Federal Flood Commission, Islamabad, Pakistan.
Handmer, J.W., 1990. Flood hazard management: British and international perspectives.
Geo Books.
Haq, N.u., Nasir, S., 2003. Water resources of Pakistan, IPRI Factfile. The Islamabad Policy
Research Institute Pakistan, Islamabad.
Hashmi, H.N., Hussain, F., Ghani, U., Mughal, S.M., 2009. Optimization of Mangla reservoir
capacity by raising dam height, 33rd International Association of Hydraulic Engineering &
Research (IAHR) Biennial Congress, Vancouver, British Columbia.
Helm, P., 1996. Integrated Risk Management for Natural and Technological Disasters, in:
Tephra (Ed.), p. 9.
Hoes, O., 2006. Aanpak wateroverlast in polders op basis van risicobeheer, CiTG. Delft
University of Technology, Delft, the Netherlands.
Hoes, O., Schuurmans, W., 2006. Flood standards or risk analyses for polder management
in the Netherlands. Irrigation and Drainage 55, S113-S119.
Hooijer, A., Klijn, F., Kwadijk, J., (eds.), P.B., 2002. Towards sustainable flood risk
management in the Rhine and Meuse River Basins. Main results of the IRMA Sponge
research project. NCR-Publication 18-2002. NCR, Delft, The Netherlands.
Hunter, P.R., Fewtrell, L., 2001. Acceptable risk, in: Fewtrell, L., Bartram, J. (Eds.), Water
Quality: Guidelines, Standards and Health. IWA Publishing, London, UK.
Hussain, E., Kim, K., Shan, J., 2009. Object–based image classification and web-mapping
techniques for flood damage assessment, ASPRS 2009 Annual Conference, Baltimore,
Maryland, USA.
IFRC, 2002. Info Bulletin No. 2/02; Pakistan: Flash floods. International Federation of Red
Cross and Red Crescent Societies.
IFRC, 2005. Pakistan: Floods. International Federation of Red Cross and Red Crescent
Societies, Islamabad.
ISDR, 2005. A review of disaster management policies and systems in Pakistan, The World
Conference on Disaster Reduction. International Strategy for Disaster Reduction, Kobe,
Japan, p. 44.
Izhar-ul-Haq, Abbas, S.T., 2008. Sedimentation of Tarbela & Mangla reservoirs, Pakistan
Engineering Congress, 70th Annual Session Proceedings Paper No 659, pp. 25-46.
J.K, V., 2001. Probabilistic design of water defense systems in The Netherlands. Reliability
Engineering & System Safety 74, 337-344.
Jarrige, J.F., 1997. Excavation at Mehrgarh - Pakistan, in: Leeuw, J.E.V.L.-D. (Ed.), South
Asian Archaeology 1975. Brill Academic Publishers, pp. 76-87.
Javed, S., Baig, M., 2005. Pasni dam bursts: Heavy rains kill 150 in Pakistan, Pakistan
Times e-paper. Times group of publicationers, Islamabad.
Jonkman, S.N., Kok, M., Ledden, M.v., Vrijling, J.K., 2009. Risk-based design of flood defence
systems: a preliminary analysis of the optimal protection level for the New Orleans
metropolitan area. Journal of Flood Risk Management 2, 170-181.
Kamal, A., 2004. Pakistan: Lai Nullah basin flood problem, Islamabad – Rawalpindi cities,
Integrated flood management. The Associated Programme on Flood Management, p. 24.
Kelly, D.J., Garvin, S.L., 2007. European Flood Strategies in Support of Resilient Buildings,
in: Ashley, R., Garvin, S., Pasche, E., Vassilopoulos, A., Zevenbergen, C. (Eds.), Advances in
Urban Flood Management. Taylor & Francis/Balkema.
Khan, A., 2007a. Preliminary damage assessment report Yemyin and Gonu cyclones
Makran coastal wetlands complex, in: Garstang, R. (Ed.). The Ministry of Environment’s
Pakistan Wetlands, Islamabad.
Khan, M.A., 2007b. Disaster preparedness for natural hazards: Current status in Pakistan.
International Centre for Integrated Mountain Development (ICIMOD) Kathmandu, Nepal.
Khosa, W.K., 2000. Kharan desert. The Federation of
Kron, A., 2007. Flood damage estimation and flood risk mapping, in: Ashley, R., Garvin, S.,
Pasche, E., Vassilopoulos, A., Zevenbergen, C. (Eds.), Advances in Urban Flood
Management. Taylor & Francis/Balkema, pp. 213-235.
Kron, W., 2002. Flood Risk = Hazard X Exposure X Vulnerability, in: al., W.B.e. (Ed.),
Second International Symposium on Flood Defence. Science Press, New York Ltd., Beijing,
China, pp. 82-97.
Kunreuther, H., 1996. Mitigating disaster losses through insurance. Journal of Risk and
Uncertainty 12, 171-187.
Lehner, B., Verdin, K., Jarvis, A., 2008. New global hydrography derived from spaceborne
elevation data. Eos, Transactions, AGU, 89(10): 93-94. http://hydrosheds.cr.usgs.gov/.
Loster, T., 1999. Flood trends and Global Change, EuroConference on Global Change and
Catastrophe Risk Management: Flood Risks in Europe. IIASA, Laxenburg, Austria.
Lyle, T.S., 2001. Non-structural flood management solutions for the lower Fraser Valley,
British columbia, School of Resources and Environmental Management. University of
London, London, p. 112.
Majeed, Z., Khan, A.I., 2008. Dam failures due to flash floods and it’s review for Mirani dam
project, Impacts of climate change on water resources management in the region,
Damascus, Syria.
Majewski, W., 2007. New approach to flood management. Publications of the Institute of
Geophysics, Polish Academy of Sciences, 401.
Medina, D., 2006. Benefit-Cost Analysis of Flood Protection Measures. Metropolitan Water
Reclamation District of Greater Chicago.
Mehlhorn, J., Feyen, F., Banovsky, I., Menzinger, I., 2005. FRAT1.0 - an example of applying
the Geomorphologic Regression approach for detailed single location flood risk
assessment, European Geosciences Union, Vienna, Austria.
Meon, G., Kleeberg, H.-B., Röttcher, K., Spanknebel, H.-G., 2006. Flood hazard map
guidelines of the German working group of the federal states on water issues (LAWA).
Bund/Länder-Arbeitsgemeinschaft, Wasser.
Merz, B., Kreibich, H., Thieken1, A., Schmidtke, R., 2004. Estimation uncertainty of direct
monetary flood damage to buildings. Natural Hazards and Earth System Sciences 4.
Merz, B., Thieken, A.H., Gocht, M., 2007. Flood risk mapping at the local scale: Concepts
and challenges, in: al, S.B.e. (Ed.), Flood Risk Management in Europe. Springer, pp. 231251.
Messner, F., Penning-Rowsell, E., Green, C., Meyer, V., Tunstall, S., Veen, A.v.d., 2007.
Evaluating flood damages: Guidance and recommendations on principles and methods,
Integrated flood risk analysis and management methodologies. FLOODsite Consortium.
Meyer, V., Haase, D., Scheuer, S., 2009. A multicriteria flood risk assessment and mapping
approach, in: Samuels, P., Huntington, S., Allsop, W., Harrop, J. (Eds.), Flood risk
management: Research and practice. Taylor & Francis Group, London.
Middelkoop, H., Asselt, M.B.A.v., Klooster, S.A.v.T., Deursen, W.P.A.v., Kwadijkd, J.C.J.,
Buitevelde, H., 2004. Perspectives on flood management in the Rhine and Meuse rivers.
River Research and Applications 20, 327-342.
Middelmann-Fernandes, M.H., 2010. Flood damage estimation beyond stage-damage
functions: An Australian example. Journal of Flood Risk Management 3, 88-96.
Mir, H., Latif, S., Mahmood, A., Hussain, S., 2010. Flood report 2010, Flood reports.
Pakistan Meteorological Department, Flood Forecasting Division, Lahore, Pakistan, p. 34.
MoE, 2003. Pakistan’s initial national communication on climate change. Ministry of
Environment, Islamabad, Pakistan, p. 92.
Moel, H.d., Alphen, J.v., Aerts, J.C.J.H., 2009. Flood maps in Europe – methods, availability
and use. Natural Hazards and Earth System Sciences 9, 289-301.
Montz, B.E., Gruntfest, E., 2002. Flash flood mitigation: recommendations for research and
applications. Global Environmental Change Part B: Environmental Hazards 4, 15-22.
MoWP, 2002a. Detailed strategy formulation, Pakistan water sector strategy. Ministry of
Water and Power, Office of the Chief Engineering Advisor, Islamabad, p. 290.
MoWP, 2002b. Medium term investment plan, Pakistan water sector strategy. Ministry of
Water and Power, Office of the Chief Engineering Advisor, Islamabad, p. 193.
MoWP, 2002c. National water sector profile, Pakistan water sector strategy. Ministry of
Water and Power, Office of the Chief Engineering Advisor, Islamabad, p. 136.
Myers, M.F., 1995. The national flood insurance program as a nonstructural mitigation
measure, U.S.- Italy Research Workshop on the Hydrometeorology, Impacts, and
Management of Extreme Floods, Perugia, Italy.
Nathwani, J.S., Lind, N.C., Pandey, M.D., 1997. Affordable safety by choice: The life quality
method. Institute for Risk Research, University of Waterloo, Waterloo, Ontario, Canada.
NDMA, 2007. National Disaster Management Authority Press Briefing - Aug 2007,
Islamabad, Pakistan.
NESPak, 2008. Second flood protection sector project – package C strengthening flood
forecasting, management and warning system. Federal Flood Commission of Pakistan,
Islamabad, p. 151.
NFRAG, 2008. Flood risk management in Australia, The National Flood Risk Advisory
Group introduces and discusses the National Flood Risk Management Guideline. The
Australian Journal of Emergency Management 23.
Park, C., 1999. Natural Hazards. Lancaster University, UK.
Pearce, D., Atkinson, G., Mourato, S., 2006. Cost-benefit analysis and the Environment:
Recent developments. Organisation for Economic Co-operation and Development.
Pelling, M., Maskrey, A., Ruiz, P., Hall, L., 2004. Reducing disaster risk: A challenge for
development. United Nations Development Programme, Bureau for Crisis Prevention and
Recovery, New York, USA.
Penning-Rowsell, E.C., Fordham, M., 1994. Floods across Europe. Hazard
assessment,modelling and management, Middlesex University Press, London, UK.
Petry, B., 2002. Coping with floods: complementarity of structural and non-structural
measure, in: al., W.B.e. (Ed.), Second International Symposium on Flood Defence. Science
Press, New York Ltd., Beijing, China.
Pielke, J.R.A., 2006. Understanding and attribution trends and projections, Workshop on
Climate Change and Disaster Losses, Hohenkammer, Germany.
Pilon, P.J., Davis, D.A., Halliday, R.A., Paulson, R., 2003. Guidelines for reducing flood losses.
Inter-Agency Secretariat of the International Strategy for Disaster Reduction (UN/ISDR),
p. 78.
Pistrika, A., Tsakiris, G., 2007. Flood risk assessment: A methodological framework,
International Symposium of EWRA: Water Resources Management, New Approaches and
Technologies, Crete, Greece.
Plate, E.J., 2002. Flood risk and flood management. Journal of Hydrology 267, 2-11.
PMD, 2009. Monsoon contingency planning update 2009. Pakistan Meteorological
Department, Islamabad.
Pottiera, N., Penning-Rowsell, E., Tunstall, S., Hubert, G., 2005. Land use and flood
protection: contrasting approaches and outcomes in France and in England and Wales.
Applied Geography 25, 1-27.
Rehman, H., Kamal, A., 2005. Indus Basin River System: Flooding and flood mitigation, 8th
International River Symposium 2005, Brisbane, Australia.
Rehman, H.U., Bhattarai, R., 2005. Modeling and socio-economic impact analysis of floods
in coastal cities under sea level rising scenarios: a case study of Karachi, Pakistan,
International Symposium on Floods in Coastal Cities under Climate Change Conditions,
Bangkok, Thailand, pp. 29-36.
Roos, A., Jonkman, B., 2006. Flood risk assessment in the Netherlands with focus on the
expected damages and loss of life, in: Schanze, J., Zeman, E., Marsalek, J. (Eds.), Flood risk
management: hazards, vulnerability and mitigation measures. Springer.
RWS, RIKZ, 2007. Atlas of Flood Maps; Examples from 19 European countries, USA and
Japan. WL, Delft Hydraulics.
Sayers, P.B., Gouldby, B.P., Simm, J.D., Meadowcroft, I., Hall, J., 2003. Risk, Performance and
Uncertainty in Flood and Coastal Defence – A Review R&D Technical Report FD2302/TR1,
Flood and Coastal Defence R&D Programme. Defra / Environment Agency.
Schanze, J., Zeman, E., Marsalek, J., 2006. Flood risk management: hazards, vulnerability
and mitigation measures. Springer.
Shah, S.M.S., Gabriel, H.F., 2002. Floods in Pakistan, Fencing Floods in South Asia: Disaster
Preparedness through Risk Communication, p. 27.
Shaikh, I.B., 2008. Water management for mitigating floods & droughts, South-Asian
Regional Workshop on Climate Change and Disaster Risk Management, Kathmando, Nepal.
Smith, D., 2004. Shortcomings in the 1% chance flood standard, Gilbert F. White National
Flood Policy Forum, 2004 Assembly, Reducing Flood Losses: Is the 1% Chance, (100-year)
Flood Standard Sufficient?, Washington D.C. USA.
Smith, D.I., 1994. Flood damage estimation - A review of urban stage-damage curves and
loss functions. Water SA 20, 231-236.
Smith, K., 1996. Environmental Hazards: Assessing Risk and Reducing Disaster. Routledge
London 2nd.
Smith, K., Ward, R., 1998. Floods: Physical processes and human impacts. John Wiley and
Sons, Chichester.
Smith, L.C., 1997. Satellite remote sensing of river inundation area, stage, and discharge: A
review. Hydrological Processes 11, 1427-1439.
Stalenberg, B., Vrijling, H., 2009. The battle of Tokyo and Dhaka against floods. Built
Environ 35, 471-491.
Tariq, M.A.U.R., 2005. Determination & Implementation of Flood Risk Zoning, Employing
Physiographic & Hydraulic Parameters for Flood Prone Areas of Chiniot Distt, Centre of
Excellence in Water Resources Engineering. University of Engineering and Technology,
Tariq, M.A.U.R., van de Giesen, N., 2010. Why Pakistan deserves generosity, The Great
Debate UK. Reuters Group Limited, London, UK.
Tariq, M.A.U.R., van de Giesen, N., Hoes, O., 2010a. Introducing a risk-based assessment
tool for sustainable flood management: A methodological framework for the evaluation of
different flood management options, International conference on sustainable water
management, 2010, Jamshoro, Pakistan.
Tariq, M.A.U.R., van de Giesen, N., Hoes, O., 2010b. Risk-based conjunctive flood
mitigation: a preliminary optimization of flood management for river Chenab, 9th
International Conference on Hydroinformatics 2010. Chemical Industry Press, Tianjin,
China, pp. 1968-1975.
Tariq, M.A.U.R., van de Giesen, N., Hoes, O., 2010c. Towards profit-loss sharing approach in
floodplain: Introducing risk-based flood insurance policy, NCR Days. The Netherlands
Center for River Research, Amsterdam, the Netherlands.
Tucci, C.E.M., 2002. Urban flooding. The Federal Public Ministry (MPF), Porto Alegre,
Tucci, C.E.M., 2007. Urban flood management. World Meteorological Organization, Porto
Alegre, RS, Brazil.
Tung, Y.-K., 2002. Risk-based design of flood defense systems, in: al., W.B.e. (Ed.), Second
International Symposium on Flood Defence. Science Press, New York Ltd., Beijing, China.
Umrani, A.P., 2001. Living with droughts, Daily Dawn. DAWN Group of Newspapers,
UN-WATER, 2005. Water hazard risks: A priority for integrated water resource
management, Building the resilience of nations and communities to disasters: elements
for a programme of action, 2005-2015. World Conference on Disaster Reduction (WCDR).
UNDHA, 1992. Internationally Agreed Glossary of Basic Terms Related to Disaster
Management. United Nations Department of Humanitarian Affairs, Geneva.
UNESCO, 2007. Balochistan and Sindh floods: Progress report September 2007. UNESCO,
Islamabad, Pakistan.
UNITAR, 2004. The Indus river basin, Training workshop on wetlands, biodiversity and
water: New tools for the ecosystem management. Kushiro International Wetland Centre,
Kushiro, Japan.
USACE, 1993. Engineering and design - Hydrologic frequency analysis. U.S. Army Corps of
Engineers, Washington DC, USA.
USACE, 1996. Engineering and design: Risk-based analysis for flood damage reduction
studies. U.S. Army Corps of Engineers, Washington DC, USA.
Van Mierlo, M.C.L.M., Vrouwenvelder, A.C.W.M., Calle, E.O.F., Vrijling, J.K., Jonkman, S.N., De
Bruijn, K.M., Weerts, A.H., 2003. Effects of river system behaviour on flood risks. Report
DC-1-211-1, Delft Cluster, Delft, The Netherlands.
Veerbeek, W., 2007. Flood induced indirect hazard loss estimation models, in: Ashley, R.,
Garvin, S., Pasche, E., Vassilopoulos, A., Zevenbergen, C. (Eds.), Advances in Urban Flood
Management. Taylor & Francis/Balkema, pp. 195-212.
Voogd, H., 2006. Combating flooding by planning: some Dutch experiences. disP 164, 5058.
Vrijling, J.K., Meijer, G.J., 1992. Probabilistic coastline position computations. Coastal
Engineering 17, 1-23.
Vrijling, J.K., van Hengel, W., Houben, R.J., 1998. Acceptable risk as a basis for design.
Reliability Engineering & System Safety 59, 141-150.
Vrijling, J.K., van Hengel, W., Houben, R.J., 2000. Response to comments by J. Ramsberg on
the paper “Acceptable risk as a basis for design”. Reliability Engineering &
System Safety 67, 211-212.
Weisz, R.N., Day, J.C., 1975. A regional planning approach to the floodplain management
problem. The Annals of Regional Science 9, 80-92.
White, G.F., Kates, R.W., Burton, I., 2001. Knowing better and losing even more: the use of
knowledge in hazards management. Global Environmental Change Part B: Environmental
Hazards 3, 81-92.
Wilke, K., 2004. Integrated flood management case study Germany: Flood management in
the Rhine and Elbe river basins. The Associated Programme on Flood Management:
Integrated flood management, The World Meteorological Organization.
Wind, H.G., Nierop, T.M., De Blois, C.J., Kok, J.L., 1999. Analysis of flood damages from the
1993 and 1995 Meuse floods. Water Resources Research 11, 3459-3465.
WMO, 2003. Mitigating risk and coping with uncertainty, World Water Assessment
Programme. World Meteorological Organization, pp. 270-290.
WMO, 2008. Tropical cyclone operational plan for the Bay of Bengal and the Arabian Sea,
Tropical cyclone programme. World Meteorological Organization, Geneva, Switzerland, p.
WWF, 2010. Hydrology of the Northern Areas, Northern Areas of Pakistan. World Wide
Fund for Nature. http://www.wwfpak.org/nap/dnap_freshwater_hydrology.php.
Yi, C.S., Lee, J.H., Shim, M.P., 2010. GIS-based distributed technique for assessing economic
loss from flood damage: pre-feasibility study for the Anyang Stream Basin in Korea.
Natural Hazards 55, 251-272.
Yoe, C., 1994. Framework for estimating national economics development benefits & other
beneficial effects of flood warning & preparedness systems. U.S. Army Corps of Engineers,
p. 116.
Young, W.J., Lam, D.C.L., Ressel, V., Wong, I.W., 2000. Development of an environmental
flows decision support system. Environmental Modelling and Software 15, 257-265.
Annex A
EAD distribution maps of existing, baseline case, fragmental dikes, and continuous dikes
(1m to 5m heights) showing the spatial distribution of EAD for the Swan River study area.
EAD distribution maps of existing, baseline case, and segmental dikes (1m to 5m heights)
showing the spatial distribution of EAD for the Swan River study area.
Annex B
EAD distribution maps of existing, baseline case, and dikes (1m to 6m heights) showing
the spatial distribution of EAD for the Chenab River study area.
Annex C
EAD distribution maps of existing, baseline, and proposed dredging, showing the spatial
distribution of EAD over the Swan River reach.
Annex D
EAD distribution maps of existing, baseline, and proposed dredging, showing the spatial
distribution of EAD over the Chenab River reach.
Annex E
Annex Figure E1: Maps, showing the expected annual net-profit of agricultural
land for the Chenab River reach against existing conditions.
Annex Figure E2: Maps, showing the expected annual net-profit of agricultural land
for the Chenab River reach against baseline case.
Annex Figure E3: Maps, showing the expected annual net-profit of agricultural land
for the Chenab River reach against 1m dike case.
Annex Figure E4: Maps, showing the expected annual net-profit of agricultural land
for the Chenab River reach against 2m dike case.
Annex Figure E5: Maps, showing the expected annual net-profit of agricultural land
for the Chenab River reach against 3m dike case.
Annex Figure E6: Maps, showing the expected annual net-profit of agricultural land
for the Chenab River reach against 4m dike case.
Annex Figure E7: Maps, showing the expected annual net-profit of agricultural land
for the Chenab River reach against 5m dike case.
Annex Figure E8: Maps, showing the expected annual net-profit of agricultural land
for the Chenab River reach against 6m dike case.
Annex F
Maps show EAD distribution of settlement areas against proposed zoning for existing,
baseline, and all dike heights in the Swan River reach.
Name: Muhammad Atiq Ur Rehman Tariq
Date of birth: June 24, 1979
Place of birth: Multan, Pakistan
E-mail: [email protected]
SMS: +92 333 57 17 47 0
Muhammad Atiq Ur Rehman Tariq started his first job as hydrological engineer in
Republic Engineering Consultants after completing his Civil Engineering in 2003. He
completed his MSc. degree with distinction in 2005. During his M.Sc., he was appointed as
research associate in a project ‘stochastic flood zoning’ in the Centre of Excellence in
Water Resources Engineering, Lahore, Pakistan due to his extra ordinary academic
performance. He then joined Pakistan’s national space agency (SUPARCO). He started his
PhD at TU Delft, the Netherlands, under the supervision of Prof. Dr. Ir. N. C. van de Giesen
in August 2007.
In the course of his PhD work, he has written several journal articles and presented his
work at conferences in Europe and Asia. At one occasion, he received the 'best research
paper award'. At another occasion, his contribution was selected for the follow-up special
issues of an international peer reviewed journal. In addition, he was invited to write an
article as technical expert for ‘Reuters’ and for a seminar lecture at Rotterdam University,
the Netherlands. He assisted several master students and fully supervised a PI project of
five engineering students.
He plays Ping-Pong and enjoys chats and discussions with friends. He prefers to attend the
company of elders to listen to their life experiences. Life is a onetime experience and we
must help all people to make their lives a good experience.
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF