This action might not be possible to undo. Are you sure you want to continue?
Assessment of Flood Control and Management Options
Final Draft: 19 June 2000
Prepared for the WCD by:
C.H. Green, D.J. Parker, S.M. Tunstall Flood Hazard Research Centre, Middlesex University
With contributions from: Luis Berga
World Commission on Dams Secretariat P.O. Box 16002, Vlaeberg, Cape Town 8018, South Africa Phone: 27 21 426 4000 Fax: 27 21 426 0036. Website: http://www.dams.org E-mail: email@example.com
This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. The views, conclusions, and recommendations contained in the working paper are not to be taken to represent the views of the Commission
World Commission on Dams Assessment of Flood Control and Management Options
This is a working paper of the World Commission on Dams - the report published herein was prepared for the Commission as part of its information gathering activity. The views, conclusions, and recommendations are not intended to represent the views of the Commission. The Commission's views, conclusions, and recommendations will be set forth in the Commission's own report. World Commission on Dams 5th Floor, Hycastle House 58 Loop Street PO Box 16002 Vlaeberg, Cape Town 8018, SOUTH AFRICA Telephone: +27 21 426 4000 Fax: +27 21 426 0036 Email: firstname.lastname@example.org http://www.dams.org
This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. The
views, conclusions, and recommendations contained in the working paper are not to be taken to represent the views of the Commission
World Commission on Dams Assessment of Flood Control and Management Options General principles
This report is about managing floods and how to identify the appropriate flood management strategy for local conditions. It is not analysis of the flood management role of dams in isolation but on how to choose the most appropriate flood management strategy in the particular local conditions. What is the most appropriate strategy will vary according to the nature of the flood and other local conditions. The recommended approach is necessarily comparative: the only way to identify the potential role of dams in flood management is in the context of the appropriate role of all other flood management options. Moreover, the most appropriate local flood management strategy will often consist of a combination of measures one of which may or may not be a dam. The most appropriate management strategy will frequently involve a mix of measures. The required logical approach is to begin by analysing the nature of the flood problem in the area, identify the available options, compare these in terms of their contribution to the society’s objectives, and finally to select the best available option. A second key principle in this report is the need for an appropriate strategy for the local problem; floods differ very widely in their nature and so do the characteristics of the floodplains affected. What is an appropriate flood management strategy in the context of one catchment will not be so in another catchment. In particular, a problem for the developing countries has been the parochialism of the approaches proposed as the solution of those countries’ problems. There has been a tendency to propose that the approaches adopted for the Rhine, Mississippi or the Thames should be applied to the Yangtze or Bramaputra although conditions are quite different in the five countries. Consequently, it is not possible to argue for or against dams in principle but it is possible to argue that there is a better option in a particular local context. What a society means by ‘better’ is then a critical question, and one that has often been addressed too narrowly in the past. It is also one that can be only answered comparatively in terms of the available options. However, in the appropriate circumstances, dams have proved to be highly effective means of reducing flood losses: whether they are the most appropriate solution in local conditions then depends on the available alternatives. Thirdly, a significant proportion of the failures of the past fifty years in water management have been institutional, and the physical solutions adopted have frequently required institutional systems to support them that have not been sustainable. Thus, the failure of those physical systems has frequently been a consequence of the failure of the supporting institutional system. At present, we would seem to be somewhat, but only somewhat, better at designing physical systems than institutional ones.
Catchment management A catchment based perspective must be taken that includes flood hazard management along with the other aspects of water and land management. Such an approach takes account of the interdependencies between functions, across geographical areas and between the use of both land and water. In most aspects of water management, runoff is the resource; in the case of floods, runoff is the problem. Thus, floods are to a greater or lesser extent an externality of land use where the area generating the runoff can be hundreds of kilometers away from the area being flooded, and frequently is in another country. In addition, the pattern of variation in the availability of runoff is often inconsistent with the pattern of human use; for water resources, the problem is then of concentrating runoff in time and space in order to satisfy the pattern of human use. Conversely, flooding is the obverse problem: runoff is too concentrated in time and space. Some parts of the world therefore suffer both problems: both flooding and an insufficiency of water at other times of the year.
This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. The
views, conclusions, and recommendations contained in the working paper are not to be taken to represent the views of the Commission
if so.g. The appropriate local flood hazard management option is dependent upon: • the nature of the local flood problem. Trying to fix rivers to a stable form often fails and is usually expensive but in some circumstances is necessary. At the same time. Consequently. Ecosystems always develop on the basis of the prevailing water regime. conclusions. In many parts of the world the floodplains have been heavily modified by past human intervention and here the dominant ecosystems will depend upon the water regime that has resulted from that intervention. less than 100 people per km2). and recommendations contained in the working paper are not to be taken to represent the views of the Commission .15 hectare of arable land per capita. particularly between those in ‘flashy’. the strategy that is appropriate for one catchment can be totally inappropriate for another catchment. the wetlands that develop in floodplains are amongst the most valuable ecological resources on the planet. will necessarily cause damage the existing ecosystems. Rivers are dynamic. and they have low intensities of economic activity for This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. Rivers are systems for transporting. what form of flood management strategy should be adopted in this area? The environmental damage caused by developing other parts of the catchment instead of the floodplain may be greater.g. modifying that water regime. they are highly urbanised (e. the appropriate economic objective is to maximise the efficiency of use of the catchment and not to minimise flood losses. Important variables in differentiating between countries (and between provinces within a country) include: the availability of arable land per capita. National differences There are no universally appropriate solutions. The two questions that should be asked are: should the intensification of development on this particular area of floodplain be encouraged? and. Taking such a catchment based approach means that terms such as ‘floodplain encroachment’ are very misleading. and those on mature river floodplains. there are no ‘off the shelf’ management strategies which invariably more appropriate than others. and • the state of development of the country. abandoning or changing those past patterns of human intervention will cause environmental damage. the degree of urbanisation. adaptive systems whose form varies as runoff and sediment loads vary over time. which is the purpose of physical flood mitigation measures. Interventions must recognise that rivers have these characteristics. Floodplains Floodplains have competitive advantages over other areas of land and therefore have always been centres for human settlements. Countries or regions will have more flood management options when: they have more than about 0. • the intensity of use of the floodplain. The views. Trends in national flood losses need not provide any guide to the success or failure of the national flood hazard management strategy adopted: it can be readily shown that efficient flood hazard management policy can be accompanied by a rise in both flood losses and the costs of flood management. • whether there is already an alleviation scheme in place. 70% of the population live in urban areas and have low population densities on average (e. population density. small and steep catchments. The nature of the flood problem can differ widely. Therefore.World Commission on Dams Assessment of Flood Control and Management Options iii In the case of flooding. the locally appropriate solution depends upon both the nature of the country and of the flood problem. depositing and eroding sediment as much as they are systems for concentrating and transporting water. and the intensity of economic activity per unit area.
for example. This might. the annual floods are expected and welcomed although the extreme floods cause death and destruction. Appraisal led design Appraisal led design means starting with identifying the objectives and then continuing to assess the identified options against these objectives throughout the iterative project cycle. rivers may be sacred spaces. including how to respond to extreme events. those up to some design standard of protection). it is unlikely that flood protection will be economically justified in rural areas although land drainage may be. environmental and distributional impacts of that project afterwards. or regarded as a waste of valuable land. It is the gain in understanding that is important rather than any numbers that emerge from the analysis. Simply proposing to adopt the approach that is appropriate in one area to another area is parochial at best and neo-colonialist at worst. After fifteen years or so. it will be more useful to look at ways of reducing the vulnerability of that population and to promote sustainable livelihoods. It is necessary to both start with the questions of ‘what is the problem? What are our objectives’ and keep asking these questions throughout the project life cycle. it is more appropriate to talk in term of ‘flood alleviation’ This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. however. those adaptations are not usually capable of coping with extreme floods. Because all strategies will fail under some conditions. These differences colour the way in which floods are managed. when the project is operating. the impact of the flood may be only a symptom of a wider problem. Rather than simply transposing the approach to flood hazard management that is appropriate for one country to another. The purpose of appraisal is to gain understanding as to what the decision involves and to communicate this understanding to all those who have an interest in the decision. in highly adapted societies.World Commission on Dams Assessment of Flood Control and Management Options iv their stage of development. that of the vulnerability of the population to a wide variety of threats. rather than simply introducing a flood alleviation strategy. Both rivers and floods are cultural constructs. it is essential to consider how all floods will be managed and not just some (e. Therefore. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . In different cultures. In developing a flood hazard management policy. The views. rather than designing the project first and then assessing the economic. into meeting some bureaucratic criteria. Traditional/indigenous societies have developed adaptations to cope with flooding that offer useful lessons. the learning that we need to transfer is the learning about what made that approach appropriate. Flood management in North Dakota with a population density of 2 people/km2 is trivially easy compared to flood management in countries such as Bangladesh and China where population densities may exceed 600 people/km2. The principles of flood hazard management Flooding should not be assumed to be the problem. Decisions concerning flood management should always be appraisal led: a preliminary assessment of the likely benefits and costs of identified options should always be made before detailed engineering and hydrological studies are made. These assessments should be progressively refined as the design process develops. the primary objective has often been transmuted into ‘completing the project’ and. In these cases. This means that it is necessary to design for failure: to consider both the flood conditions under which a particular strategy will fail and the mode of failure. the appraisal process must be transparent and open to public involvement. once all direct and indirect agricultural subsidies have been removed. in highly urbanised countries with industrial scale farming.g. include introducing soft credit and empowering the members of that society. valued for the environmental and resources they support. conclusions. In general.
as well as a low peak level. provide benefits in the form of replenishing soil fertility. again. For agriculture. It is also desirable to avoid discontinuities in the appropriate form of adaptation by those at risk e. the conditional probability of death should a flood occur is low as compared to other hazards. But. The ideal flood to manage is slow rising. control of the rate at which it is taken up and then the rate at which the stored volume is released are important as well as the amount that is stored. including whether a floodplain should be developed and the form of the flood management policy to do adopt. In the UK. Desynchronising the peak flows from different tributaries is often an important part of flood management in order to reduce the peak flow level on the main stem of the river. the moisture in the root zone must be within certain limits. With storage. Storage. Thus. we mean simultaneously both the correct and the just decision. Different societies however differ in what they take to be the overall goal of societal decision making. and then to identify a strategy to manage the risk in those areas affected. it was usual to speak of identifying some over-arching ‘public’ or ‘national’ interest whilst in other countries the choice may be defined solely in terms of a conflict of private interests. The appraisal process In making societal decisions. and often both at different times of the year.World Commission on Dams Assessment of Flood Control and Management Options v or ‘flood mitigation’ than of ‘flood protection’ or ‘flood control’. and proportions of. losses from floods are typically much less important than the losses from inadequate land drainage and floods can. conclusions. whether this storage be in the soil or surface water storage. has a central role in controlling both conversions. This may involve land drainage and/or irrigation. precipitation is converted into runoff and then runoff is turned into river flow downstream. but do not always. The definition of sustainable development agreed at the Rio Conference specifies that two defining conditions of the right process are the involvement of the public at all levels of decision making and the recognition of the role of women. we seek to try to take the ‘right’ decision. it is more appropriate to think in terms of water level management than of flood management: to maximise production.g. floods have a high probability of occurrence. Slowing the time to peak is particularly important when flood warning lead times are short and the risk to life is significant. has been taken by the runoff by earlier event precipitation events or the start of the individual event then it will have little effect in mitigating the current flood. For dryland farming. however. One consequence of this approach is that a flood management strategy will typically involve a combination of different approaches rather than reliance on a single one. flashy. storage is an important means of desynchronising peak tributary flows. we also mean both achieving the right outcome and doing so by the right decision process. the antecedent conditions are important: if the soil has been saturated by previous precipitation or surface water storage. with a long time to peak. there are contexts in which this conditional probability is high and it is essential in developing any flood hazard management strategy to identify those conditions where there will be a high risk to life. Consequently. both are problems on small. Consequently. The views. Some societies also characterise the aim of the decision This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. By ‘right’ decision. Compared to other hazards. in Continental Europe it is common to speak of ‘social’ or ‘national’ solidarity and Islamic societies take account of the duties owed to other people and to other species. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . controlling soil water levels through drainage will normally be a high priority than protecting against out of bank flows from watercourses. steep catchments. points at which a particular strategy of coping with flooding suddenly becomes a counter-productive form of adaptation for a more extreme flood. In general. natural or artificial. In all forms of storage. it is desirable to slow the rates at which. There is no ‘right’ answer but the ideological power of economics can result in the neglect of the reality that there is more to life than economic efficiency.
Therefore. the most useful approach is to involve the public early on in the identification of both important issues and possible options. where farming is industrialised. or. there is therefore no decision to be made if we are certain what option is the best option: we are uncertain either because of the complexity of the choice or because there is an inherent conflict between the options. identifying the potential role of dams in flood hazard management requires simultaneously identifying the potential roles of all other flood hazard management options. and not what precisely the future will be. However. The strategy identified through the appraisal process can be no better than the best of the options compared. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . knowing that we will make mistakes and must seek to learn from them. Rather than pretending that we can know the future. or even less differentiated expectations about the future. The views. Uncertainty is the lack of differentiation and can exist either or both between outcomes and/or the probabilities of each outcome. Other combinations are possible and we are seldom so lucky as to be making a decision under risk. The adaptive management approach implies that we should have a preference for options that involve resilient natural systems. In a real sense. managed retreat and resettlement of those currently living on the floodplain should normally be amongst the options considered. It is this understanding that is important. The reason for using these approaches is to inform the decision process. it is critical that an adequate range of options be considered if the best option is to be found amongst them. In turn. a choice only exists if the available options are mutually exclusive and we can choose only one or another option. when we cannot differentiate between the probabilities attached to each option but only between the outcomes. the form of uncertainty that is important is decision uncertainty: uncertainty about what to do. multi-criteria analysis).g. None of the available techniques is without drawbacks and limitations. decisions as to the flood hazard management policy to be adopted should be appraisal led: the different available strategies should be compared against the societal objectives brought to the choice. benefit-cost analysis. In urbanised countries. in terms of their consequences. In any event. Decisions under risk are those where we can differentiate between the probabilities and the outcomes associated with each option. to clarify the nature of the choice that must be made. we should instead approach all decisions as being part of a learning process.World Commission on Dams Assessment of Flood Control and Management Options vi process to be to achieve a consensus whilst others define societal choices in a much more adversial way. Whilst some rules have been proposed so as to increase the likelihood that the best option is included amongst those considered. Consequently. Choices are also necessarily be made under uncertainty. it is only necessary to decide what is the best of the options available. because they lie in the future. forgiving artificial systems or enhance the coping capacity of individuals and communities. Choice is also inherently about conflict. Outcome uncertainties must not be allowed to obscure decision uncertainty and the latter can be easier to manage. an adaptive management approach should be adopted. conclusions. In order to promote this understanding. the appraisal method needs to be transparent and all assumptions should be explicit. Choices are made between alternatives and appraisals are therefore necessarily comparative. This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. equivalently. It is dangerous to treat a decision under uncertainty as simply being one under risk and it has been argued then when confronted with uncertainty. the advantages and disadvantages of one option exist only relative to those of another option. The state of being uncertain can then be defined as a state of rational doubt as to what to do. particularly as to the nature of the conflicts involved in the choice. the consequences of all the options are shrouded in uncertainty. then we are making a decision under uncertainty. A number of different techniques exist to aid in the appraisal process (e. The reasons why the alternatives are mutually exclusive vary and an important aspect of the appraisal process is to identify the nature of the conflict that is involved.
particularly in urban areas. The simplest form of model is a flood hazard map. for example. its affects on water availability in general and not just on flood runoffs should be considered. the nature of the flood hazard must be modelled. For physical interventions. lines should never be used both because there will be major uncertainties and because a line implies a significant difference between conditions on the two sides of the line. Flood management options Before a choice can be made. In modifying the challenge presented by the flood. environmental changes. All physical interventions will have consequences both for the water and sediment regimes. frequently as a result of lack of funds. Therefore. the designer should provide a maintenance schedule setting out the maintenance actions required. the most appropriate management strategy will involve a combination of approaches. In particular. will be more efficient than flood proofing. it is essential to consider how that intervention will be maintained and how the necessary funds for maintenance will be generated.g. Constant natural assets are those where individual sites may be lost provide that the total stock remains constant. This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. because this is what they are intended to do. In the case of physical interventions. frozen or saturated ground) must be considered. the map should not be limited to the outline of the 100 year return period because to do so violates the principle that all floods must be managed and not just some. Very vulnerable areas of population should also be shown. The views. but not necessarily will be. at some intensity of development a dike. Source control can be considered as a form of storage in the soil or via the soil. Source control should normally be considered. and with a low peak level.World Commission on Dams Assessment of Flood Control and Management Options vii Consideration should be given to incorporating environmental impacts into the decision process by introducing the concepts of critical natural capital and constant natural assets as constraints. the costs of a dike to protect an area are a function of the length of bank involved whilst the costs of flood proofing individual properties in that area are a function of the number of properties in that area. an example of this approach is the ‘no net loss’ and ‘wetland banking’ policy adopted in the USA. Usually. Similarly. In both cases. Critical natural capital covers those environmental assets that are too important to be sacrificed except in the most extreme circumstances. Shading should be used throughout. and therefore for the environment. Thus. we are seeking to modify the flood so that it is the easiest type with which to cope: slow rising. A by-product of institutional changes may be. with a long time to peak. On such maps should be indicated the critical characteristics of the flood hazard. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . All interventions should be expected to have some impact on existing settlement patterns and result in some resettlement. conclusions. floodways and storage areas. there are major economies of scale for collective flood alleviation over individual flood mitigation. such as the depth and velocity of flooding. pre-flood conditions (e. The flood management options available can be characterised as those that reduce the challenge and those that enhance the capacity of individuals and society to cope with the flood. A common cause of failure of both physical and institutional interventions is a lack of maintenance. there tend to be economies of scale in the provision of physical structures: a large area of storage being lower in cost than a number of small areas having the same total capacity. Native species should normally be used. In assessing the likely effectiveness of source controls. A similar approach is also necessary for institutional strategies. for example. The form of intervention can also be categorised into those that involve a physical intervention versus those that involve an institutional change. When considering afforestation. their frequency and cost in the same way that a bill of quantities will be part of the contract documents.
these flash floods are amongst those that present the greatest risks to life. investment is focused on improving the technologies required to improve flood forecasts. with the storage being designed to control the more extreme flood flows. Dams. Institutional maintenance is essential if the management of a flood is to be effective. There is a necessary trade-off between warning lead time and forecast reliability. Typically. Warning and evacuation depend for their success on prior emergency planning. An ‘all hazards’ emergency plan should be prepared and the emergency plan should treated as the development of a network of individuals and organisations rather than as a document. Detention basins can be similarly useful in lowland areas to cap the peak of flow floods. by way of dams and detention basins. continuing rehearsals and cooperation between the parties involved is a requirement of successful emergency planning. It is essential to start by finding out the user needs of those at risk and also of the different organisations who will be involved in translating the forecast into a warning and then in disseminating the warning. in the appropriate circumstances. in very flashy catchments in particular it is unwise to rely upon the reliability of a formal flood warning system. Control of both the rates of inflow to the storage and of the outflow has significant advantages in that this avoids storage capacity being taken up by the rising arm of the flood. This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. planning constraints are unlikely to work. and particularly of capping the peak flows of the more extreme floods. Dikes are most likely to be appropriate for floodplains that are already intensely used. Check dams have the added advantage in some areas of enabling the recharge of groundwater although their primary role is generally for erosion control. Where land is under development pressure. however. especially from informal development. then providing incentives for development to be undertaken elsewhere will probably work better than simply trying to stop development on the floodplain which may merely result in corruption. conclusions. is most likely to be appropriate on small. as may be engineered wetlands. Land use control: if intensified development on a particular floodplain is undesirable. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . It is necessary that the information contained in them be disseminated to the public at risk and they should be based on realistic expectations of how the public will behave in a flood. The population should generally evacuate or be evacuated prior to the flood and not attempt to sit out the flood in their dwelling. the weakness of flood warning systems is usually found in the process of converting that forecast into a warning and getting that warning to those who need it in time for it to be useful for them. Storage and dikes can be a particularly useful combination of approaches. Emergency planning is about co-ordination and co-operation between institutions rather than a schedule of actions. a reliable and effective warning system is consequently difficult to achieve when warnings must be based on predicted rainfall because the time between the rainfall and the flood is otherwise too short to disseminate a warning. and have a history of interventions to promote flood alleviation. Flood warnings are always necessary but dissemination takes time. such as urban areas and rural areas in countries that are not urbanised. flashy catchments and in complex river systems where it is necessary to avoid the flood crests from different tributaries coinciding. Unfortunately. storage is most usefully applied to remove the peak of the flood. Flood proofing or house raising are most likely to be appropriate where development intensities are low and properties are scattered.World Commission on Dams Assessment of Flood Control and Management Options viii Surface water storage. The views. can be a highly effective way of reducing downstream flood losses. Emergency plans are essential for areas lying behind dikes and below dams.
Holistic management of the catchment will increasingly result in multi-functional options being adopted. however. it is usual for government to provide some level of compensation to victims of disasters. > 2 metres.World Commission on Dams Assessment of Flood Control and Management Options ix Evacuation is essential where the buildings or other features do not provide a safe place of refuge during a flood. In general. Different institutions have funding restrictions that reflect their functional and geographical boundaries. Since decisions are made and implemented by institutions. evacuation may be upward (e. In general. it is essential to establish ways of co-ordinating across boundaries. The different disciplines are one aspect of institutions. conclusions. There are a number of barriers to be overcome to achievement of this objective. There is an apparent tension between promoting public involvement which tends to imply the devolution of decision making to the lowest possible level and holistic management of the catchment which pulls towards a multi-functional institution that covers the entire catchment. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . Except in partnership with government. flood insurance is an adjunct to a flood hazard management policy rather than being a tool for implementing such a policy. There are different possible forms of such a partnership and these can have the effect of transferring some or all of the burden of compensation on to those who buy insurance. For outward evacuation to be successful. not least to promote accountability. The views. not masonry or concrete framed). catchment management authorities should be responsible downwards rather than upwards. Depending upon circumstances. and ones that can be maintained. The principle of managing the catchment as a whole means that multi-functional options are becoming increasingly important. institutional design is critical to the success of a flood hazard management policy. In most countries. when a specialist is appointed to examine This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. the insurance industry has treated floods as an uninsurable risk. Reducing the barriers to achieving sustainable flood management Sustainable flood management strategies are those appropriate to local conditions. it must be planned in advance and the population concerned must know what to do in a flood emergency. A general funding protocol should be negotiated instead. agreeing the funding of a multi-functional approach on a case by case basis with the different institutions can be time consuming where it is possible at all. Thus. Different countries construe the nature of society in different ways. flood velocities are high (> 2 metres/sec).g. Where this is done then can be advantages to sub-contracting the assessment and distribution of compensation to the insurance industry because of its expertise in this area. The nature of the discipline defines what it does and equally what is expected of that discipline.g. An artificial wetland can. they should have a board of directors made up of representatives from the organisations. for example. the government will always be the reinsurer of last resort and for that reason it is advisable that the nature of this public-private partnership be formalised. into a flood refuge) or outward. To minimise these problems. Because the scope for redefining institutional boundaries is typically limited for historical. authorities and others living and using the catchment rather than being responsible solely to higher levels of government. cultural or other reasons. Outward evacuation will generally be necessary where the depths of water are significant (e. or buildings are flimsy (e. One boundary that can prove a major hindrance to achieving such holistic solutions is single functional funding. All institutions necessarily have geographical and functional boundaries. increase biodiversity and remove pollutants from the water. as determined through public involvement. provide flood storage.g.
society. engineering organisations are increasingly becoming multi-disciplinary and they have arguably progressed further in this direction than organisations deriving from other disciplines. or approach. engineers play an enabling role. Indeed. When giving aid. We learn at least as much from our mistakes as our successes. The population who benefit should normally bear the costs of operating and maintaining the strategy.World Commission on Dams Assessment of Flood Control and Management Options x a problem. In practice. conclusions. This means more than making data available. has implied the expected solution. the extent of its contribution varying according to the relative wealth of the area in question. is necessary for long run success. Developing countries should seek to learn from the mistakes made by the more developed countries rather than to replicate them. If engineers became engineers in order to build things. increasing the likelihood that the most appropriate strategy will be identified. and to studying flooding from a disciplinary orientation. Hypothecated taxes for this purpose are more transparent This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. Conversely. The start is to ask the public what are the questions they want answered. pure scientists want to study it. For equity reasons. by deeming that specialist’s discipline to be appropriate to the problem. At the same time. But societies are changing rapidly to be more pluralistic and the nature of public service is changing with it. Local ownership of flood alleviation strategies. better flood management will only emerge from better research. The understanding of the nature of the choices that must be made in flood management is enriched and widened by involvement of all interested parties. Public involvement does not mean allowing the system to be captured by those with the most power. Engineering training needs to be orientated towards this wider role. Similarly. engineers are more problem orientated than most disciplines and this is an aspect of engineering training that should be encouraged. Here. In the long run. However. society expects engineers to build things. the problem is that the disciplinary focus tends orientate academics towards developing the discipline in the eyes of their peers. Study tours of the developed world should. it also means providing the public with the means to convert raw data into useful information. to the problem. Now. influence or social competency. The views. engineers have to listen to the different publics and then assist those publics in achieving their goals. be concerned more with identifying what went wrong in the developed countries rather than looking for solutions that can be imported. engineers are the bridge between decision makers and pure scientists between whom otherwise there is frequently a major gap: decisions makers have to do something. alternative viewpoints can enrich the problem space. rather than developing inter-disciplinary perspectives that speak to the needs of the decision makers. engineers sought to determine what the public needed and then provided it. Both are forms of neocolonialism. governments have been known to put the interests of their national companies ahead of those of the recipients. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . the objective should be to skip the generation of mistakes made by the developed countries so as to leapfrog them. The interests of others must be protected and integrated into the decision process. In earlier generations. therefore. and the options involved. Democraticising information is a prerequisite of effective public involvement. and similarly NGOs to put he interests of their members ahead of the sustainable livelihoods of those who live with floods. Water engineers have had a long history of public service. all those who seek involvement have to answer for their legitimacy: what right are they entitled for their voice to be heard in a decision? One claim for legitimacy is that their concern is with the sustainable livelihood of those who will be affected by the decision to be made. But the institutional boundary in Universities is between different disciplines and here again the problem is to determine how to build co-operation across these boundaries. it will often be appropriate for the central government to contribute towards the capital costs of the strategy adopted.
flood control dams reduced flood flows by half. are determined by local conditions. the advantages and disadvantages of each available option being compared. The forms of storage of storage that are possible. the choice of flood management strategy must be made on a comparative basis. The California Flood Emergency Action Team (1999) report on the 1997 floods claimed that on a number of major river systems. the case for a dam is strengthened. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . The role of dams in flood management In the appropriate circumstances. and • the floodplain is heavily developed. such as irrigation or power generation. This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. • multiple tributaries contribute to the local flood problem and it is important to prevent the flood crests from the different tributaries being synchronised. the storage provided behind dams can result in a substantial reduction in flood losses. to reiterate. the US Army Corps of Engineers annual reports to the US Congress routinely claim very substantial reductions in flood losses as a result of the storage provided in dams. the latter being particularly helpful in flood management. • Small scale ‘warping’ and ‘catch dams’ can frequently form a useful component of a strategy to control soil erosion and sediment movement in areas where the sediment loads generated would otherwise be considerable. Storage in soil or in surface waters will reduce total flood flows and controlled storage reduces the peak flood flow. • the ratio of flow in an extreme flood to the flow of the annual flood is high. if any. Where the reservoir created can also be used for other purposes. A dam is most likely to form part of the appropriate management strategy when: • the major part of the runoff comes from a small.World Commission on Dams Assessment of Flood Control and Management Options xi than paying these costs out of general revenue and a locally accountable organisation is more appropriate than a more centralised body. But. steep catchment immediately above the area at risk. • the time to concentration is short. conclusions. The views. Positioning the storage in relation to the areas at risk and the timing of storage uptake and release are both important to the choice of the form of storage to adopt and the impact of the storage on the flood.
...... 4...................................................................................................................1 Indigenous Flood Adaptations .....27 Risks to Life and Health ...............45 4..................................42 Managing all Floods and Not Just Some.....................................2 Flood Control and Defence..........................................................................World Commission on Dams Assessment of Flood Control and Management Options xii CONTENTS 1................................4 Environmental consequences of flooding.................................................4.................4...........5 1.5 1.................................... INTRODUCTION. 3..........40 3.............................................3 Offshore Impacts of Flooding .................................................................48 This is a working paper prepared for the World Commission on Dams as part of its information gathering activities..........35 3.3 Livestock Benefits and Losses .....21 3.2.....................................................2 Land Drainage ... The views..................................................................................1 1...............................................................................................................2.............................................................................4 Integrated catchment management and the environment ......................................................2....2 The Four Generations of Flood Hazard Management ....... and recommendations contained in the working paper are not to be taken to represent the views of the Commission ...............................................4......................................34 3.....................................2................................................9 1............................33 3..............................1 Measures of the global flood problem ....37 3..........................................................................39 3...1 Flood Losses............................................................................4 The Evolution of Dams in Flood Hazard Management............................1 Unregulated Rivers.......................3 Urban impacts ........................... 2.............................................7 1...............................................1 The Nature Of Flood Problems..11 2...................................................................8 1...................43 4..........1.......36 3.................................................... conclusions..........................3 Non-Structural Approaches..................................2............2.........................1 2........1 1............................1 FLOOD MANAGEMENT OPTIONS .......4......................................2 Agricultural Benefits and Costs...................................................................................................................................................................................2......1 Alternative definitions of vulnerability.........................................................................................................................................3 1.....................9 Management .......... and ‘Efficient’ Rivers .................................3 Identifying the threat ..........................................1 THE BENEFITS AND COSTS OF FLOODS .............................7 1...................................................................................................................30 3.............................. FLOOD GENERATION AND FLOODPLAIN ..............2 Managed and Regulated Rivers.46 4............2 What is Vulnerability and Who is Vulnerable ? ....................................40 3...............................................2 THE NATURE OF FLOODS.........4 Holistic Approaches ........................................................................................40 4..............3 The Cultural Construction of Floods and Flood Hazard ........................2.........39 3.....................18 Flood Production..18 The Use Of Floodplains ....................
..120 Public involvement ......7 The environmental effects of the different options...........9 THE ROLE OF DAMS IN FLOOD HAZARD MANAGEMENT ..........................4..........................2 Storage.............................................................................................................121 8.................................................6 Matching the Solution to the Problem ........................................... and recommendations contained in the working paper are not to be taken to represent the views of the Commission ...................................................................................................................83 4.......................4...................50 4..........4......117 Maintenance.....................................................................................................2 7.4 Carrying the flow ..................................................................70 4..97 MAKING THE ‘RIGHT’ DECISION? .................................62 4..........4 7..5 Enhancing coping capacity.........................................................5 7...............................................................................................................................117 Financing..........82 4..............120 Legitimising voices .............119 Research ...4.................................................................72 4......... The views..3 7..........1 Controlling runoff .........4 Compensation..............................................................................60 4................63 4..................1 Emergency Planning and management ...................................4.....................77 4.................................................. 7..........63 4.....................3 Slowing the flood wave....................................................................................................................................................................................................................................4.................................................................................................................1 7.............................76 4...................................World Commission on Dams Assessment of Flood Control and Management Options xiii 4.............5.........................................................................4...........122 REFERENCES AND BIBLIOGRAPHY ..........................................................................................................124 This is a working paper prepared for the World Commission on Dams as part of its information gathering activities.................................................................. 9......................5 Flood Insurance ...........3 Evacuation....5....................75 4...........................114 Professional roles......113 Institutional barriers.................................................................................................................................8 7......5................................................117 Exporting/importing failure....5 Separating the people and the threat ..................................................7 7..............................................102 BARRIERS TO SUSTAINABLE FLOOD MANAGEMENT .......................................................................................5............ conclusions.........................................6 7....................2 Flood Forecasting and Warning ................................5................................................................................................ 7.............88 4...88 5....................5........................................................................................... CONCLUSIONS AND RECOMMENDATIONS.................6 Help and Counselling................................................................119 Corruption ...........6 Economies of scale in flood alleviation ............................84 4...............58 4..4 Reducing the Challenge ................................................................................. 6............
World Commission on Dams Assessment of Flood Control and Management Options 1 1. the strategy adopted must consider how all floods are to be managed and not just some. These definitions suggest many different types of floods based initially upon the prime causal agent (Table 1). The views. morrains Snowmelt Icemelt Flooding during freeze-up Flooding by ice breakup Mudfloods Coastal/sea/tidal floods Dam This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. At the same time. Jokulhlaup Spate floods Riverine Riverine (also called ice-jam floods) Floods with high sediment content Induced by volcanic activity Storm surge (tropical or temperate induced) Ocean swell floods Tsunamis (induced by geological process) Dam-break flood Dam overtopping Failure of natural dams e.g. Introduction Floodplains are areas where either there are ecologically important wetlands. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . conclusions. and are also areas that have competitive advantages for human settlement.1 The Nature Of Flood Problems An economically-worded but broad definition of a flood is offered by Ward (1978): ‘A flood is a body of water which rises to overflow land which is not normally submerged’. Table 1 AGENT Rainfall Types of flood DETAILS AND EXAMPLES Riverine or non-riverine Slow-onset or flash flood Convectional/frontal/orographic Torrential rainfall floods Riverine Overland flow Glacial meltwater (rise in air temperature) Glacial meltwater (geothermal heat source) . Resolving the potential conflict between ecological value and human use is consequently a major issue in determining the most appropriate flood hazard management strategy. 1.e. or were such areas in the past.g. The likely outcome is then that the strategy adopted will consist of a mix of options.
The concept of ‘vulnerability’ is central to an understanding of flood hazards and to the definition of the appropriate management response and is explored further below. conclusions. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. and channel characteristics (e. A crucial part of the concept of a flood hazard is that the interface between floods and people. and the antecedent conditions as well as the characteristics of the stream networks characteristics (e.e. To this must be coupled the nature of the terrain which generates the runoff (e. channel length).g. A flood is not hazardous unless humans are somehow affected. Floods have a number of measurable characteristics. This is taken further by Hewitt (1983) when he states that a hazard refers to the potential for damage that exists only in the presence of a vulnerable human population. volume) or magnitude. cyclones and typhoons. particularly when this rainfall occurs on a steep catchment. including flood depth or ‘stage’. frequency (usually estimated as a return period or recurrence interval). The views. and frontal rainfall may be influenced by topography to generate orographic rainfall floods. Central to the concept of a ‘flood hazard’ is the notion that hazards are an ever-present conditions which periodically lead to harm. This ever-present condition is likely to vary in intensity since many types of floods are characterised by seasonality – during the flood season the hazard is more intense. rising sea levels. A ‘flood hazard’ is the threat to life. storage capacity. _____________________________________________________________________ Floods are part of the dynamic variation of the hydrological cycle. the basic causes of which are climatological. velocity. property and other valued resources presented by a body of water which might rise and flow over land that is not normally submerged. extent and seasonality. Many of the most catastrophic floods are then associated with the intense rainfalls that result from with hurricanes.World Commission on Dams Assessment of Flood Control and Management Options Sewer/urban drain flood Storm discharge to sewers and drains exceeds capacity Burst water mains Breaches in canals 2 Water distribution failure Rising water tables (high groundwater tables) Land subsidence. channel roughness and shape) (Ward 1978). soil type and vegetation cover). For example: convectional rainfall may generate flash floods.g. duration. reductions in abstractions from aquifers Note: Flood types are not necessarily mutually exclusive. The latter tend to be related to characteristics of the terrain so that steep catchments are associated with narrow rivers with low storage capacity. discharge (i. geology.g.
The views. Swiss Re (2000) give an estimate of 55. al. Floods are ranked slightly lower when the definition of disaster is broadened to include civil strife. the same data reveal that floods affect the lives of more people (an average of 65.World Commission on Dams Assessment of Flood Control and Management Options 3 1. Neatly presented data sets may appear to be consistent and precise but the underlying data collection processes may be unreliable. The following data and the conclusions drawn from them should therefore be treated with caution. The average annual number of people made homeless by floods between 1972 and 1996 was also the highest for any disaster type (3.874) was the second highest for any type of disaster. These estimates are corroborated by more recent data from Munich Reinsurance for the period 1986-1995 (United Nations. Data from IFCRCS (1998: 142-45) reveal that between 1972 and 1996.360 as the number of deaths caused by flooding in 1999. This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. All available estimates of the regional distribution of loss of life by disaster type reveal that (a) all disasters and (b) flood disasters. conclusions. 1992). averaging 89 per year over the period 1988-97 (US Army Corps of Engineers 1998a) but the per capita and total economic losses appear very high because of high standards of living and high values at risk. increased reporting of events over time and many other factors. the annual average number of people injured by floods (21. Munich Reinsurance (1997) report that fifty-five per cent of deaths (367. drought. In North America the number of flood-related deaths is comparatively low.000 people) were caused by flooding.36 million). Flood disasters are among the world’s most frequent and damaging types of disaster (IFRCRS: International Federation of Red Cross and Red Crescent Societies) 1998: 147). 1992). and that both have a markedly skewed distribution with by far the highest reported deaths occurring in Asia. al. recent estimates of average annual flood damage for the 1987-1996 period reveal that the entire region’s flood damage losses now exceed those of the Americas and Europe (IFCRCS 1998: 147). are ubiquitous. Glickman et.000 having died in the mudfloods and landslides in Venezuela alone. In addition. During the latter half of the twentieth century floods were the most common type of geophysical disaster. famine and high winds killed more people than floods. including drought and famine. but in the 1986-1995 period floods appear to have caused more deaths than any other geophysical disaster type according to Munich Reinsurance (UNDHA 1997:6). Data on flood effects is also prone to political manipulation in order to secure aid. However. drought and famine (which are all excluded from data of Glickman et.87 million per annum between 1972 and 1996) over the same period than any other disaster type. Earthquakes and tropical cyclones kill more people than any other geophysical disaster type. al. For this period. earthquake.1 Measures of the global flood problem Measuring the world’s flood problem presents many difficulties because of deficiencies in the quality of statistics. Because of economic growth in parts of Asia. some 20.1. generating over thirty per cent of all disasters between 1945 and 1986 (Table 2) (Glickman et. (1992) indicate that globally. flood disasters are about the third most harmful form of geophysical disaster in terms of loss of life. Department of Humanitarian Affairs (UNDHA) b1997: 6). and recommendations contained in the working paper are not to be taken to represent the views of the Commission .
may not be particularly significant because. the escalation of flood losses in the United States. and the monetary value of property losses and other economic losses has been steadily increasing. It is difficult to identify trends in the frequency and impacts of disasters such as floods. This is a working paper prepared for the World Commission on Dams as part of its information gathering activities.072) 131 223 275 191 (189) 27 7 1. Governments and individual may alternatively have to borrow heavily to fund these replacements and repairs. Data source: Glickman et.. some estimates (Parker 1996) have been produced for a small number of countries revealing widely varying proportions of total country populations which are flood-prone.5 per cent in France. Unfortunately. These are 3. However. conclusions. 4.World Commission on Dams Assessment of Flood Control and Management Options Table 2 Global loss of life by geophysical event. there is no comprehensive global or even national data bank to draw upon for such data. The United States is probably the country where the most thorough analyses have been performed. al.198 (388) 40 3 6. Two trends appear in the reported flood loss data for the United States: loss of life has been reduced significantly during the twentieth century and appears to have been constant for a number of years. is subject to a number of well-known defects (Mishan 1967). as possibly elsewhere.033) 295 157 1.8 per cent in the United States. Individual floods can cause significant losses to the economic capacity of a country. Another way of scaling the world’s flood problem is to examine estimates of the number of people and properties located (or exposed) in flood-prone areas.343 (1. for every credit. relative to the Gross National Product of that country.849 (837) Note: numbers in parenthesis exclude the three worst disasters. Gross Domestic Product (GDP). (1992). not least because the standard national accounting model. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . economic losses from floods appear to have held constant over the past fifty years (Wooley 1986).907 (1. the costs of replacing damaged or destroyed infrastructure may absorb the resources that would otherwise be available for economic or social development.264) 25 6 2.494 271 85 40 1.8 per cent in the United Kingdom. 1945-1986 Deaths (in thousands) 244 791 (291) 28 5 4 4 Type of disaster Meteorological Flood Tropical cyclone Other storm Heat wave Cold wave Geological Earthquake Volcanic eruption Tsunami Other Landslide Fire Total Number of disasters 395 272 (271) 212 23 15 Deaths per disaster 618 2. The effect is to minimise the apparent effect of a flood on the national economy in most instances.053) 1. The extent of these losses is difficult to measure. over 50 per cent in the Netherlands and 80 per cent in Bangladesh. The views. 9.267 (1. However. there is an equal and opposite debit and vice versa.272 (2. The basis of the GDP measure is equivalent to a form of double entry book-keeping. the construction of 500 schools will appear in the same way whether these are to extend educational provision or to replace schools destroyed in a flood.
We need also to constantly re-assess the management approaches adopted and to transfer knowledge from one country to another. In Bangladesh not only are some dwellings deliberately constructed on higher ground. it has been argued (Wittfogel 1957) that the large scale works of the Middle East and Asia were the consequence of strong central governments.2 The Four Generations of Flood Hazard Management Over time.e. we have gained technical understanding about the nature of floods and the means of managing whilst at the same time our perceptions of floods and rivers have changed. The knowledge that we need to transfer is the understanding which resulted in that approach being identified as being appropriate and not necessarily the approach itself. dwellings and possessions may be lost. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . Historically.World Commission on Dams Assessment of Flood Control and Management Options 5 1. The base floor of the dwelling construction typical of this area was set upon cypress pilings (or stilts) sunk into the silt deposited by spring and summer flooding (Laska and Wetmore 2000). In the early stages of the evolution of flood management strategies these local adaptations may be the only. The views. through urbanisation and economic growth) indigenous approaches are eroded (Chan and Parker 1996). it is a form of neo-colonialism. The Cajun or Acadian population that joined the native American communities in the coastal swamps and marshes of Louisiana in the eighteenth century also adapted their dwellings to floods. In addition. houses are constructed on stilts to raise them above anticipated flood levels and the use of small boats is common.2. in river estuaries and along coastlines. Conversely. 1. traditional villages) of Malaysia. Too often however rather than knowledge transfer taking place. four different approaches to flood hazard management have succeeded each other. Flood embankments were part of these indigenous approaches. These are just a few examples of the numerous indigenous adaptations to flood problems which are used by those who have lived and worked for generations in flood-prone areas. approaches to flood risk management have changed over time. form of flood management. conclusions. These indigenous approaches are often relatively effective in rural areas as long as exceptional floods are not encountered when numerous lives. the use of flood-tolerant rice crops and the use of boat-craft instead of roads since these and bridges may be washed away. Such actions started quite early and were widespread over Western Europe (Wagret 1967). this is a form of parochialism. This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. the approach developed in the context of one country has simply been transposed to another country.e. Similar adaptations to floods may be found throughout South-East Asia along rivers. but some are dismantled in times flood and moved to the top of earthen flood embankments. for example. in the floodplain kampungs (i. communities banding together to construct dike systems. local adaptations to make them more resilient to flood hazards and disasters. For example. Experience suggests that as modernisation takes place (i. partly in order to convert wetlands to arable land. At best.1 Indigenous Flood Adaptations Communities which have occupied flood-prone areas for many generations have typically developed usually small-scale. these systems being constructed and maintained through a system of Common Property Resource management (Ostrom 1990). the agricultural economy is adapted to flooding through. Thus. or the dominant. at worse.
The estimates performed immediately after the flooding based on the reports of local flood prevention committees specified the global damages and losses countrywide at c $US 2. i. The flood’s most tragic effects were 54 fatalities. (5 . conclusions. agriculture and forest industry – c $US 0. Also affected were 4. flood levels reached in the upper Oder were estimated as millennium flood-waves: they were up to 2-3 m higher than the absolute maximum ever recorded. The existing flood defence infrastructure turned out to be inadequate. This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. As a result.22 July) in the south-western parts of Poland and northern parts of the Czech Republic.000 employed over 5 persons. In the Vistula basin the flood had a more restricted character.3 billion (in 1997 prices). the precipitation was more limited and the swelling of the waters less violent.5 billion • Losses incurred by companies and firms excl. it was impossible to effectively reduce the flood wave. The inadequate technical condition of the flood levees. While the area flooded then was twice as large in 1997. Comparisons with the 1934 flooding indicate a changing pattern of flood damages. the number of damages bridges 38 times larger.2 times larger.000 budget units and gminas with 680.5 – 2. resulted in numerous dike failures. The flood also revealed shortages in the technical equipment of the civil defence and emergency services and the inadequacy of the flood prevention action. Some 152.5 billion or 2. Unit runoffs reached 500 l/s per square kilometre. The rain gauges. in these regions.8 billion • Losses n household equipment – c $US 0.6 percent of GDP. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . Water swelling was mainly caused by both several-day periods of intensive precipitation.World Commission on Dams Assessment of Flood Control and Management Options The 1997 Flood in Poland 6 The July flood of 1997 was one of the largest that has been recorded in Poland.9 July and 18 .85 billion • Agriculture and forest industry – $US 0. The most intensive rainfall occurred in the upper parts of the Oder (Odra) River basin. registered precipitation at the level of 300 – 600 mm during several days. and the length of damaged roads 134 times larger in 1997. 2 to 3 times more than the total average monthly amount of precipitation in July within those gages (the largest precipitation took place in the Czech Republic). the number of flooded buildings was 3. Particularly large losses occurred in urbanised areas due partially to uncontrolled construction in potentially flood-prone areas in the recent years.000 companies were located in the flooded or partially flooded area of which 9. Due to the limited number of flood surcharge reservoirs.e. The views.7 – 0. The report mostly focused on direct losses and did not consider the value of facility equipment. flood result forecasting and disseminating warnings are concerned. Damage due to flooding was reported in ca 20 percent of the communities (gminas). The breakdown of losses is as follows: • Losses in assets of budget units and gminas (communities) destruction of streets and public roads and water systems – c $US 1. Slightly later estimates performed by GUS based on surveys set global losses at c $US 3.4 billion The long-term influence of flood damages caused to the economy is quite difficult to estimate. especially as far as communications.000 household (according to the poll conducted by the main statistic office (GUS) in early 1988). most of which were constructed back in the 19th century.
Purseglove 1988). as in Hungary.World Commission on Dams Assessment of Flood Control and Management Options 7 The flood was an impulse for upgrading to a broadly defined flood prevention system for the country. by associations of local land owners (Vituki 1998). large-scale engineering) measures but many of the works of this time where undertaken by Provincial or Local government. the ‘non-structural’ options frequently came to be offered as an alternative to the traditional engineering solutions. Unfortunately. 1964). constructing flood relief channels and sometimes constructing a series of flood control dams.2 Flood Control and Defence. It provides for the creation of a modern national monitoring and forecasting system. Madej Pawe Roman Konieczny Institute of Meteorology and Water Management Krakow. perverse impacts on downstream areas (making their flood problems worse). The benefits of such strategies have been large during particular periods. structural approaches have a number of disadvantages. Whereas the second phase defined the problem as the rivers. was the ‘engineering’. Poland 1. flood embankments may be only partly effective in exceptional floods (i. and the solution This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. It was often marked by state promotion of structural (i. and preparing for future floods. on the one hand. they may be overtopped or breached). Although these were originally proposed as part of an integrated strategy for the good management of floodplains (White 1945. The emphasis in this approach was to control the river and to prevent floodwater entering communities located in flood-prone areas. The language used reflects this struggle to make rivers efficient servants of human purposes: floods were to be ‘controlled’ and ‘defences’ to be prepared against floods. A Flood Recovery Project based on the World Bank loan was initiated in 1998 aiming at eliminating damage from the 1998 flood. and flood control may only address a part of the problems which cause flood disasters (i. on the other. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . 1. including that flood control structures may encourage further floodplain development.2. There was a strong emphasis on building flood embankments designed and constructed to engineering standards. Much of the history of flood mitigation in the United States during the nineteenth and twentieth centuries (until the late-1960s) was based upon this strategy as the US Corps of Engineers struggled to control great rivers such as the Mississippi. structural approaches may have adverse or damaging environmental consequences (Brookes 1988.e. or.2.e. conclusions. For example.3 Non-Structural Approaches The third phase was the advocacy of ‘non-structural’ approaches. the modernisation of the lagging economies of the Tennessee river basin during the middle part of the twentieth century was driven by a strategy to control river flooding and soil erosion by building a series of large dams which had other benefits such as generating electricity for rural electrification programmes.e. The views. development of a flood prevention strategy for the two main drainage basins in southern Poland and activating local communities for flood preparations with particular focus on non-structural measures. The philosophy was strongly rational: rivers being ‘trained’ or ‘improved’ to become efficient and to stop floods interfering with human activity. flood control does not address people’s vulnerability to flood hazards). characteristic of the late nineteenth and most the twentieth century. scientifically rational approach to river management. and ‘Efficient’ Rivers The second generation approach.
Rather than engineering the rivers so as to be efficient. conclusions. and the recurrence of exceptional and highly damaging floods.2. they should bear the consequences of their choice. There may also be an effort to encourage purchase of flood insurance. and ‘flood mitigation’. Thus. Flood-proofing (which is a planned approach to modifying buildings to make them more resilient to flooding) builds upon indigenous flood adaptation approaches (see 1.World Commission on Dams Assessment of Flood Control and Management Options 8 being to keep the rivers away from the public. holistic approaches may be seen as a return to this original idea. This third wave was characterised by the same optimism as the second wave of approaches. if they insisted upon occupying the floodplain. it cannot be extended to cover the control of development on flood plains. and of ‘flood hazard management’ or ‘flood risk management’. This emerging approach talks in terms of ‘flood alleviation’. traditional engineering approaches and there was insufficient recognition of the difficulties of making the approaches actually work. or of the preconditions necessary for them to work. Not uncommonly non-structural analyses implied that people should not be on the floodplain in the first place. the development involves not just economic development but also human development.1 above) and this may be promoted. It also includes the emphasis on intra-generational.4 Holistic Approaches The original idea behind the non-structural approach was expressed in the terms of ‘coping with floods’ or ‘living with floods’. and to be a replacement for. planning controls may be proposed to prevent the spread of communities on to the floodplains. Thus. In extreme cases entire communities or parts of communities have been moved from flood-prone to flood-free land. as may improved flood forecasting and warning schemes to allow people and property to be evacuated from a flood-prone area in advance of a flood. non-structural approaches were argued on the basis that the public should be kept away from the rivers. for example. Mileti 1999. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . The vision is one of deliberately designing and planning communities that are adapted to floods in a variety of ways. Critical evaluation of the successes and failures of the non-structural strategies. so strongly advocated in the United States. United Nations 1992) is one of the drivers of this emergent approach. and that the ‘non-structural’ model of flood management. Planning controls that to seek to ensure that new buildings are flood proofed against some design standard flood are also typical of the non-structuralist approach. for example. has led to a variety of strategies (and not a single strategy) that is based upon a more holistic approach to addressing the basic causes of floods and flood disasters. Non-structural approaches were generally assumed to offer an alternative. The views. Myers and Passerini 2000). many structural and non-structural strategies have failed to be sufficiently effective. Non-structural approaches include small-scale ‘structural’ modifications of individual buildings (designed-in or retrofitted adaptations) and measures designed to move people away from floods. It requires rethinking because This is a working paper prepared for the World Commission on Dams as part of its information gathering activities.2. and. requires rethinking (Changnon 1996. It is now becoming clear that in the United States. being a concept that only started to be applied in the middle of this century and until a land use planning system has been effectively established. but also the interrelationships between land and water. in the early part of this century. the requirement to think about the catchment as a whole. not just in terms of the geographical and functional interdependencies involved. the approach centred upon making people behave. In particular. The concept of sustainable development (ACC/ISGWR 1992. as well inter-generational. It is also more critical and less optimistic than either of the previous two waves. 1. Secondly. although some past investments in flood control structures proved to be wise. rather than in terms of flood control. including increasing public involvement in decision making. equity often missed when the Brundtland definition of sustainable development (World Commission on Environment and Development 1987) is quoted. only engineering options were available: land use planning.
In Bangladesh. conclusions. The second example stems from analysis of the causes of floods particularly in the less-developed regions of the world where lack of access to resources. although technically an extreme event is simply a phenomenon which differs substantially from the mean. • on the wise use of floodplains and coastal zones (not necessarily moving out of them).World Commission on Dams Assessment of Flood Control and Management Options 9 of its inadequacies in the United States and because of its poor applicability to many other world regions.al. exacerbating the lasting effects of these events (Blaikie et. and • on addressing problems of intra and inter-generational equity (e. Urban development results in permeable natural surfaces being replaced by artificial impermeable ones so that runoff accumulates more completely and rapidly in stream and river channels rather than infiltrating into the soil and percolating into groundwater. ‘flooding’ is a normal environmental condition to which people have become adapted and which This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. Failure to grasp the importance of these variations when designing flood management strategies is likely to lead to adverse consequences. dealing with poverty and lack of access to resources as a means of addressing flood vulnerability). and recommendations contained in the working paper are not to be taken to represent the views of the Commission . for example. in contrast to many parts of the world. a division is drawn between the ‘flood’ that is the normal seasonal inundation of a floodplain to which traditional settlement and land use is well adapted. The first is the advocacy of source control approaches to flood management (Gardiner 1994). This emergent approach leads to an emphasis: • on holistic catchment and coastal zone management. inappropriateness and failure. 1. Approximately 80 per cent of the land area of Bangladesh comprises river floodplain. Here. In rapidly urbanising catchments in particular. This analysis leads to the view that floods and other hazards need to be addressed systemically through stimulating social and economic development rather than only through flood alleviation schemes: a ‘sustainable livelihoods’ approach.g. A whole-catchment planning approach is required. The views. ‘Flood’ is a relative concept and varies amongst cultures and individuals in its perception and understanding. and poverty generate a vulnerability amongst people which makes them particularly vulnerable to floods. • on improving local capacities to respond. and ‘floods’ which represent abnormal or unwanted flooding that causes loss (Paul 1984. • on valuing and preserving the best of indigenous adaptations. developing a mechanism for extending low-cost loans to those of low economic status may be the kind of strategy which holds the best promise for the future in terms of making people more resilient to the effects of floods. The approach adopted should be stakeholder driven. 1994). Two examples illustrate the emerging holistic analysis. • on empowering local communities to make choices about land development and flood alleviation). 1987). promoting flood disaster resilience (Handmer and Dovers 1996). Thus. to education and to decision-makers.3 The Cultural Construction of Floods and Flood Hazard Management Both floods and flood hazard management are culturally constructed phenomena: people’s understanding of them derives from their unique environmental and cultural conditions. Sources of flood flows are thus everywhere where development is occurring and it is here that runoff retarding and flood storage measures are required. The latter might be referred to as an ‘extreme event’. is development – not just in the floodplain but everywhere that it is occurring in the catchment. the driving force behind growing flood discharges and worsening flood hazards. • of reducing the impacts of humans on the environment.
whilst others perceive them as sacred places or places to be reserved for wilderness and the preservation of ecosystems in which human intervention should be limited. The roles of the individual and state in flood hazard reduction represents a further example of cultural construction of flood hazard management.g. What would be called a flood in Britain or Germany is not perceived of as a flood in Bangladesh. and the values they associate with them. planning authorities. but like earthquakes. a hazard must be foreseeable and it must also be possible to do something. Blame follows from rationality. whereas ‘floods’ are threatening. and whilst both the causes of floods and the reasons why a flood has not occurred recently are frequently misperceived (Green et al 1991). people in Bangladesh frequently blame floods on the actions of those in India. as in the cases of the Netherlands and China. Although there are strong commonalties in the cases cited of dwelling design generated by the nature of floods. the Constitution stating that ‘the inhabitability of the country and the protection of the protection and improvement of the environment are public tasks’ (Huisman et al 1998). even in Britain. probably an over-optimistic belief. Rivers are also cultural constructs because people’s perceptions of them. some organisation or individual must be seen as having had the responsibility to do some thing but have failed to take such action for one or another reason. Tunstall and Fordham 1991). flood hazard management is also a function of cultural and environmental factors (Green. people seek causes. an ancient principle underlying English flood legislation is that responsibility for drainage of land and avoidance of flooding rests first and foremost with the individual riparian owner. ‘Normal’ flooding is therefore a resource for Bangladeshi cultivators and one to which they look forward. public authorities will increasingly be held to account for their actions or inactions and judged against the public’s expectations of how they ought to behave. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. To go from cause to blame does require two additional stages: there must first be a belief that the flood could have been controlled. there is generally a belief. The views. For example. a flood is a relative concept because individuals living in various parts of a floodplain have different understandings of what constitutes a flood to them. environmental protection groups) and the public and/or national interest. conclusions. In the Netherlands where flooding is a strategic threat to the nation. commerce. As the discussion in 1. and a complex history of case law has accumulated to help settle disputes between riparian owners. a belief as to the cause of a flood typically then leads on to blaming someone or some organisation for failure to act. In other cases. although this is relatively uncommon (Islam 1997). These examples illustrate the varying bases for governmental intervention in flood hazard management which are found around the globe.1. Public perceptions about the causes of floods and flood disasters vary enormously around the world. An increasing expectation that some thing will be done is also a corollary of increasing public involvement. Only in the lowest-lying areas did there emerge historic organisations to address flood problems (Darby 1983). Some may perceive of rivers as places to dispose of waste. that it is possible to control floods or their consequences. building contractors. vary enormously from culture to culture and within the same culture. In some cases floods may be perceived of as ‘an act of god’ or ‘the will of Allah’. To be controllable. floods are generally regarded as foreseeable. those living in floodplains for generations develop indigenous adaptations which are their own cultural adaptations to floods. and in the case of China it is about organising national solidarity against a threat to the nation. Secondly. damaging and entirely unwanted. governmental intervention is more of a strategic affair. Unlike earthquakes. But Gateley (1973) points out that. the basis for government decision-making is seeking a balance between the actions of various stakeholders in floods and floodplains (e.2. China and Nepal where they see the floods as having originated. there is more of a collective and consensual approach. illustrates. residents of flood-prone homes. Thus. as is partly the case in England and Wales.World Commission on Dams Assessment of Flood Control and Management Options 10 provides much needed soil moisture for crop growth on which people depend. prevented or at least mitigated. developers. In some cases. More generally. floods are blamed on someone or some organisation not doing their job.
a weir is a This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. Weirs in stream also create some storage capacity although usually primarily created for other purposes such maintaining navigations. climate change) • Constitution reserves power to set land and building controls to individual States.World Commission on Dams Assessment of Flood Control and Management Options Table 3 11 The comparative evolution of flood management: determinants of flood management policies The Netherlands • Very high population density • Very high intensity of economic activity • Relatively small amount of arable land per capita • Landscapes and ecosystems are largely the consequence of past human activity • Whilst politically the Dutch approach is one of consensus decision making. a wide mix of strategies. some State constitutions reserve that power to local government • Settled only in the C19th • Low proportion of land is at risk of flooding • Historic reliance on dikes as engineering approach to flood alleviation • Historically.4 The Evolution of Dams in Flood Hazard Management It is useful to separate the issues of what a dam does.g. acid rain. Balancing lakes are artificial lakes connected in-line with the watercourse. from that of how it is done. In a sense. in an extreme flood. flood alleviation funded through Federal government The strategy of flood management adopted in a country therefore reflects the culture of that country and local flood conditions. have been adopted • Works are undertaken by all levels of government but predominantly by local government USA • Very low population density • Very low intensity of economic activity • Very large amount of arable land per capita • Large areas of country are essentially untouched by human activity except by large scale changes (e. therefore. 1. either the lake level will go on rising or the natural throttle on the discharge will fail catastrophically. Lakes and wetlands both regulate flow through storage. the outflow channel is not dimensioned so as in principle to be able to discharge the extreme flood. conclusions. including the earliest known example of flood warning. flood alleviation funded through local Common Property approach China • Very high population density • Quite high intensity of economic activity • Very low amount of arable land per capita • Landscapes and ecosystems are largely the consequences of past human activity • Central government can require land and building controls but has difficulties in enforcing such requirements • Floodplains settled in C9-C13th or earlier • High proportion of usable land is at risk of flooding • Over last two thousand years. as is the case with ‘yokulhlaup’. Unlike reservoirs. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . Table 3 summarises the geographical and cultural factors that have shaped flood management practices in three countries. lakes by having a naturally formed channel which to a greater or lesser extent limits the discharge capacity. central government can require land and building controls • Settled in medieval period onward • Most of the country is reclaimed from sea and floodplains • Historic reliance on dikes as engineering approach to flood alleviation • Historically. The views. this strategy is thus shaped by history. regulating variations in river flow through storage. Thus. the failure of a glacier releasing the stored up flood water behind it.
the means of creating a reservoir. although the formal definition of large dams specifically excludes dikes. the ‘average’ dam is not a large dam.375 dams are included as large dams in the World Register of Dams out of a total of 75. Otherwise they have much in common in terms of the nature of the construction and the consequences of failure. the important feature of a dam is the quantity of storage provided. The views.000 small dams world-wide. Thus. In the United States.855 are given in World Register of Dams 1988 (ICOLD 1999). China has declared 24. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . Table 4 Number of dams constructed at least in part for flood alleviation purposes in Japan (Source: River Bureau 1985) National/Public corporation Prefectural Government multi-purpose flood control completed 73 189 143 46 262 under construction 89 205 147 58 294 Total 162 394 290 104 556 Table 5 Dams and flood management in China (Source: Tong and Xiaogan 1997) This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. this gives 11 more single purpose flood alleviation dams as either completed or under construction than are given in the ICOLD tabulation where Japan only registers dams over 30 metres high (ICOLD 1999).187 dams in the United States. Similarly. A reservoir is then an in-line form of storage which creates a significantly greater head than a weir. For flood alleviation. Some dikes even exceed the simpler definition of a ‘large dam’ being more than 15 metres high. The purpose of detention basins is also to regulate flows through storage but these are typically off-line: many being dry or semi-dry until flow is diverted into them. or levee or flood embankment. Table 4 summarises the use of dams for flood control purposes in Japan as at March 1985. As a structure. unlike dams whose primary purpose is hydropower where the head achieved is also important. 6. of which details of 1. the distinguishing feature of a dam being that is perpendicular to the flow of the river whereas a dike.World Commission on Dams Assessment of Flood Control and Management Options 12 low head dam. is parallel to that flow. conclusions. Thus. The Environment Defence Fund (1999) estimates that there are around 800. The available data on the use of dams for flood alleviation purposes is not wholly consistent and is also partial.671 large dams to be in operation.
conclusions.900 13. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . this was some 70% of the PMF and had an estimated return period of 500 years.061 37. generally. a water conservation dam. priorities change over time: dams which were constructed to reflect one set of needs have their operating rules changed. 8% of dams are reported as having flood alleviation as one of their purposes (Lecornu 1998). Most dams which have flood alleviation as a purpose are declared as multi-purpose dams although there are national differences. The highest proportion is in Germany which declares nearly 56% of all dams as including flood alleviation as one of the functions of a dam. the flood inflow to the Francisco Morazon (El-Cajon) dam was estimated as 9.900 The proportion of dams which are declared to ICOLD as having some flood alleviation function varies from 0% (Sweden. Conversely. USCOLD. Alternatively.700 295. then there is a fairly natural tendency to ask whether it cannot be put to some other use whilst waiting for the 100 year return period flood to occur. Nearly 50% of dams in the USA which have a flood alleviation function are declared as having that as their single function and nearly 80% in Brazil.800m3/sec. Equally. amongst other summaries of the reductions in losses from individual floods as a result of the operation of dams. Thus. cites the Lexington Dam. and may be enlarged.400 m3 Song hua jiang River Huai he River Yangtze Zhu River jiang River 13 Total 187 113 345 2. as opposed to the zero reported to ICOLD. or disadvantageous. Thus. Nor is data necessarily consistent within countries : in India. reporting differences is likely to be the main explanation for these national differences and it makes international comparisons unreliable It is not surprising that a low proportion of dams globally are reported as having flood alleviation as their sole role: once a reservoir is in existence. Romania. Globally. Again. India and Norway) through to over 40% in the cases of Argentina. SANDRP (2000) cite 13 dams as having been declared by the Central Board of Irrigation and Power as having flood control as one their purposes.World Commission on Dams Assessment of Flood Control and Management Options River characteristics Liao he Hai he Huang River River he River (Yello w River) Large reservoirs 126 85 140 165 Reservoir capacity (million 26.500 54. There is no obvious pattern to be observed in terms of the proportions of dams that are declared as having a flood alleviation function between different countries and it is likely that a significant proportion of the differences are no more than the result of national differences in the manner of recording the purposes of multi-functional dams in particular.400 38.200 m3/sec. dams on the tributaries of the Yangtze built primarily for hydropower are now being examined to see to what extent they can be used for flood alleviation purposes as well (Daoxi and Siping 1999). impact on downstream flooding. when a different priority emerges. the river passes through a narrow This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. any dam will have some advantageous. as having reduced flood losses in 1952-53 by more than $US 3 million (UCOLD 1999). China reports only 5% have flood alleviation as their sole function and Turkey 3%. Mirstkhoullava (1994) described how the settlements along the River Dnieper are protected by Dnieper cascade of dams: flood peaks are reduced by 2050% and the dams have a capacity to store 25 km2 of flood waters. Saad (1999) has demonstrated the impact of the Aswan High Dam on flooding in the Nile valley and Berga (1999) provides a general summary of the role of dams in flood mitigation. the explanation may lie in a difference in engineering cultures in different countries. The reservoir stored some 1500 million m3 and the maximum discharge was 1. The views. In addition.000 25. this will be beneficial (Bergstrom and Lindstrom 1999). Downstream of the dam.000 100. During Hurricane Mitch. Japan and the Czech Republic. In any event the consequence is to make international comparisons unreliable.
Both examples are typical of Japan in that both topographical and social reasons.164 m3/sec during the flood of 1990. this has already been noted as having a very high ratio of flood flow to average annual flow.000 years. and the peak inflow to the dam reached 1309 m3/second but peak discharge from the dam was limited to 715 m3/second. The views. The Committee on Dams and Floods of ICOLD (2000) detail the effectiveness of a number of dams in reducing flood flows. The hydrographs for the inflow and discharge volumes also illustrate the very rapid rate of rise of the river. above that gorge is an intensely populated alluvial plain. means that leaving open wide floodplains both to carry This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. Tokyo is on the floodplain of the River Tone and flood alleviation works were first undertaken in C16th. flood damages were considerable. review comment). peak rainfall reached 60 mm/hour and the peak inflow to the Matsubara dam reached 2911 m3/sec. the flood flow more than quadruples and rises by a factor of approximately sixty over little more than sixteen hours. when the inflow to the Chungju dam in Korea reached 22. in a period of about six hours. including bypass channels.000 m3/sec estimated for the 100 year return period event. having been progressively enlarged over several centuries. during a seasonal rainfront in 1982. the flood occurred at probably the worst possible time: the early morning when most people are asleep and consequently the likely effectiveness of most forms of warning is very low. if the flow downstream had not be attenuated by the reservoir. Severe damages were experienced in the flood of 1947 and one element of the revised flood management strategy adopted in 1949 for the catchment was the construction of the Shimokubo dam. it is difficult to construct a single large dam and instead a cascade of dams is used to control the peak flow at critical points downstream.000 m3/sec which is somewhat less than the flow of 14. The peak discharge from the dam was limited to 1044 m3/sec. However. In terms of issuing flood warnings. unusually high. with the response to the second wave of rainfall appearing to be almost instantaneous. and the high population density. The very low amount of arable land available in Japan (Table 7). are intended to the limit peak flows from the 200 year return period flood to the equivalent of those from the 50 year return period flood. Similarly. the series of dams including the Shimokubo. rainfall over three days on the catchment reached 319 mm. The ratio of the 1. in the case of the Tone. so reducing the height of the dikes or width of the floodplain required to safely pass the flood flow. on the Chikugo river system. with one hectare of arable land having to provide food for eight times the global average number of people. the peak outflow was less than 12. They include the Tone catchment in Japan. then the losses downstream would have been catastrophic (Palmieri. nearly 90 times annual mean flow. the dams are part of a comprehensive flood management strategy with both dikes and channel improvements. and recommendations contained in the working paper are not to be taken to represent the views of the Commission .000 year flood flow to the 100 year flood flow in the case of the Chungju dam is also. During typhoon #10 in 1982. Thus. considered to have a return period of around 1.World Commission on Dams Assessment of Flood Control and Management Options 14 gorge with limited flow capacity. In addition. Put figure from ICOLD report here Figure 1 ___________________________________________________________________________ Similarly. in world terms. Upstream of the dam. the very high ratios of flood flows to mean annual flow make sole reliance on the later approaches extremely difficult since either very high dikes or a very wide floodplain are required. conclusions.
World Commission on Dams Assessment of Flood Control and Management Options
floods and provide some natural storage is not an option. Equally, high dikes are wide dikes and both the cost and land take increase rapidly with height, although Japan in now constructing some super dikes, dikes wide enough for buildings to be constructed on the crest and rear berm (Rivers Bureau 2000). Limiting peak flows through both dams and the retention basins that are also used is consequently a logical option. In general, the Committee does not provide estimates of the reduction in flood losses yielded by the dams discussed, and there are a number of stages necessary before a calculated flood flow can be converted into flood losses. Few if any of these works were constructed after a formal benefit-cost analysis and detailed estimates of the reductions in flood losses are not generally available. In addition, it was noted earlier (Section 2.2) that in economic terms, the objective is not to reduce flood losses but to increase the efficiency of use of the catchment. However, the estimates of the US Army Corps of Engineers of the reductions in flood losses resulting from the flood storage reservoirs and basins, and the dike system, in 1993 Mississippi and Missouri floods are cited (US Army Corps of Engineers 1995): Dams 3.6 7.4 11.0 levees 3.9 4.1 8.0 All (billion US$) 8 11.5 19.1
The 1993 flood was an extreme event, with an estimated return period of 500 years in some areas and at least the 100 year return period event in 88 out of the 154 streamflow gauging stations (US Army Corps of Engineers 1995). The Committee on Dams and Floods also points out that the US Army Corps of Engineers estimates that the combination of upstream storage and other works reduced the flood peak at St Louis by two metres, with the result that flood crest was one metre below the top of the floodwall that protected the business district. Throughout this report, the necessity of comparing options has been stressed. Thus, one of the more interesting figures on the effectiveness of dams is given in the US Army Corps of Engineers assessment of flood management options for the Mississippi system. Table 6 summarises the calculated impacts on flood levels of a number of management options which would either decrease runoff or increase storage. The figures are for the Omaha District (chosen here because the figures are most complete and generally show the greatest effects). This particular example shows that the reservoirs had a significantly greater effect than the alternatives tested, one advantage of reservoirs or detention basins being that the use of the storage can be timed to have the greatest effect upon the flood peak.
Table 6 Flood management options for the Upper Mississippi and Lower Missouri Rivers Omaha district Remove existing agricultural levees (and -3 to –4.5 ft continued agricultural usage) Remove existing agricultural levees (and -0.7 to –2.3 ft reversion to natural vegetation) -0.4 to –1.4 ft Set back levees
This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. The
views, conclusions, and recommendations contained in the working paper are not to be taken to represent the views of the Commission
World Commission on Dams Assessment of Flood Control and Management Options Remove existing reservoirs +3 to +9 ft flood duration increased from 0 to 60 at Sioux City, 1 to 67 days at Omaha, 25 to 80 days at Nebraska City -0.8 to –1.4 ft
Runoff reduction by 10% (Note: St Paul District estimated that would require conversion of 2.5 million acres to wetland to provide storage equivalent to 10% reduction in runoff) (Source: US Army Corps of Engineers 1995)
Following a catastrophic flood in 1913, which resulted in the loss of 360 lives, local residents in the Miami Valley in Ohio pressed for improved flood alleviation measures. New legislation permitted the creation of a Conservancy with the power to raise taxes to undertake flood alleviation works. Prior to constructing a series of detention dams and other works, the Conservancy District assessed the benefits and costs of the plan for some 77,000 individual tracts of land along some 195 kilometers of river. One purpose of this exercise was to set equitable taxes but a second requirement was that the benefit-cost ratio should exceed three. Following completion in 1929, the dams have been used over 1,000 times. This is a particularly interesting case because the works were funded by the local community. More detailed data is available for the Cerrillos Dam in Puerto Rico. This is a 98 metre high dam that protects the city of Ponce at the mouth of the Bucana River (Committee on Dams and Floods 2000). The Bucana is a classic flashy river with a long narrow catchment with an area of 81 km2, mainly of tropical vegetation, and an average slope of 55 meters per kilometer. Consequently, flood velocities in Ponce have been measured as up to 2.4 metres/second, and thus pose a significant risk to life (Section 3.1). The Cerrillos Dam is designed to provide adequate flood storage for flood with a peak inflow of 1342 m3/sec, resulting from 440 mm of runoff over the catchment, and a inflow hydrograph volume of 29x106m3, for which it provides storage for 20x106m3 – rather greater than the 100 year return period flood volume. The value of the property at risk of flooding in Ponce was estimated as US$600 million. The Committee on Dams and Floods reports that when Hurricane Hortense struck Puerto Rico in 1996, the partially complete works reduced flood losses by US$130 million. Hurricane Georges resulted in daily rainfall amounts of 550mm and 450 mm on the 21st and 22nd September, with a peak inflow to the Cerrillos Dam of 1506 m3/sec, estimated to be the 170-year runoff event (US Army Corps of Engineers 1999). The Corps estimates that the reduction in flood losses was approximately US$320 million, and that otherwise the depth of flood water in the centre of Ponce would have been 1.4 metres. As in the Japanese examples, the introduction of storage was coupled with channel improvements and other control works in Ponce. The US Army Corps of Engineering reports annually (US Army Corps of Engineers 1998, 1999) to the US Congress on both the estimated flood losses and the reductions in flood losses that resulted from existing flood alleviation works. Since the emphasis in flood alleviation is on catchment management, the reductions in losses are not generally attributed to dams versus levees. In addition, as in the Mississippi flood of 1993, the operation rules for the dams were often varied to take account of the conditions experienced so as to aid flood fighting in the areas downstream. But, the reports do ascribe significant reductions in flood losses to the operation of dams; for example, the main stem
This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. The
views, conclusions, and recommendations contained in the working paper are not to be taken to represent the views of the Commission
World Commission on Dams Assessment of Flood Control and Management Options
reservoirs on the Missouri reducing losses by $US 5.2 billion during the early spring flood in 1997 (US Army Corps of Engineers 1998). A useful statistic to have would have been the relative importance of dams within the flood management strategies of different countries and the extent to which there are systematic differences between countries in the use of dams, although such differences would probably largely reflect differences in topography. It is not possible to break down the ICOLD figures to see whether there have been changing patterns of construction of dams for flood alleviation purposes over time. The caveats made above about probable differences in national reporting patterns would also make using anything except a single nation’s figures of dubious reliability. Figure 2 illustrates the variation over the years in the number of all types of dams built in different countries. How these plots should be interpreted can be argued; one possible interpretation is as a wave of development which must decline as sites for dams are exhausted. Why any such wave exists is also a phenomena whose meaning could be contested; one interpretation being that dams have been regarded as a symbol of modernisation and independence – an argument that has been proposed in the case of Spain.
Reported trends over time in the number of dams built
% of all dams by decade built
60 50 40 % 30 20 10 0 <1900 19101919 19301939 19501959 19701979 19901999 USA UK Spain Japan India Turkey Italy France Korea
This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. The
views, conclusions, and recommendations contained in the working paper are not to be taken to represent the views of the Commission
a high proportion of precipitation is converted to runoff. The Yellow River basin in northern China is simultaneously an area of enormous flood risk and of water shortage (World Bank 1997). Particularly in arid zones. In floods the problem is the runoff and the rate at which it is concentrated both over space but more especially in time. How much precipitation is converted to runoff and how quickly depends on the form of the impermeability of the catchment and how quickly the water is conveyed off the surface. It is rarely possible to manage a catchment with only a single purpose. in the Crocodile river basin in South Africa. Flood Generation And Floodplain Development 2. Runoff also depends upon the capacity of the surface either to retain moisture or to allow it to enter groundwater. runoff is the resource: rivers simply concentrate and convey the water. In naturally impermeable surfaces such as rocks and clays. not infrequently the additional runoff generated exceeds the potable water required to sustain them. Afforestation reduces all runoff and not just the runoff from the events that cause flooding. conclusions. in small. they pose particularly severe problems. the ideal pattern of runoff is one that shows little variation from year to year and is concentrated in the growing seasons. Both the intensity of precipitation and the areal extent of the precipitation event are important. such as flood alleviation.World Commission on Dams Assessment of Flood Control and Management Options 18 2. steep rocky catchments. Urban areas are largely impermeable and cities can be seen as mechanisms for harvesting rainfall. The form and nature of the catchment determines how much of the precipitation is concentrated into the river and how quickly. Surface water drainage systems in urban areas are usually designed to get ride of runoff as quickly by discharging it to the nearest water course. rainfall can also be intense and North Africa. variations not only within years but also between years can be substantial (Smith 1999). the capacity of an industrial forest to capture rainfall is reducing the availability of water not only to the demand areas downstream but also to the Krueger National Park. Because tropical storms typically resulting in intense rainfall over a large area. Land drainage systems in agricultural and forest areas can have the same effect with the result that the flood peak is both higher and occurs more rapidly than under natural conditions. partly because rainfall is both highly variable and concentrated in the summer months. afforestation has other benefits. Runoff also depends upon previous conditions: when the soil is already saturated. terracing and afforestation to control soil erosion (Dixon et al 1994) have also increased water scarcity (World Bank 1997). Sandy soils thus retain little moisture but highly organic soils can absorb large quantities of precipitation and thus reduce runoff. The standard water availability indices (Falkenmark and Widstrand 1992) of runoff per capita can therefore be seriously misleading as indicators of the availability of water as a resource. In general terms. Topography is also critical. is frozen or is baked hard by the sun. for example. Many of the most catastrophic floods that have occurred have been the result of rain falling on land whose capacity to absorb water has already been taken up by earlier falls of rain. Similarly. notably carbon fixing and in terms of biodiversity. in mind. a high proportion of precipitation is converted to runoff. In Shanxi Province of China. The Nature Of Floods. Human activity can change not only the proportion of precipitation that is converted to runoff but also the rate at which that runoff reaches the water courses. if carried out using native species. flash floods can be produced within a few minutes of the rainfall event with water velocities of up to 15 metres a second so that such flood This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . For human purposes. The views. has experienced very severe flash flooding.1 Flood Production In water resource management. In arid areas. On the other hand. the pattern of human demand for water is generally subject to less variation in time over the year than the variation in climate.
there were 8 flood crests as the rain front moved west up the Yangtze (Daoxi and Siping 1999). it can be seen as an externality of land use in an area perhaps hundreds of kilometres from the area affected by flooding. both ‘check’ and ‘warping’ dams can have a role to play in capturing sediment before it enters the main river as well as in creating fertile areas on valley bottoms (Development Alternatives 1999. one reason why flooding is increasing in Bangkok is because of excessive groundwater abstraction has lead to falls in ground level. Some rivers. Leung 1999. their capacity to carry the sediment load depending on their flow. Therefore. New South Wales 1999). rather there is likely to be a succession of runoff events as the rain front.World Commission on Dams Assessment of Flood Control and Management Options 19 events are quite capable of moving large boulders as well as uprooting trees. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . a large river is fed by a multiplicity of tributaries draining sub-catchments. rivers also concentrate and transport sediment. This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. Part of the management strategy for the Yangtze therefore is to seek to prevent the flood peaks generated by the different tributaries and the main stem coinciding in time at any point on the main river. are almost a mudflow rather than a river. The time taken for runoff to reach a point on the main stem depends upon which tributary generated the runoff and from what event. Thus. In addition to runoff. It is unlikely that a rainstorm will simultaneously effect the entire catchment. such as the Yellow River in China. for example. At the same time. flooding is only a symptom of a problem rather than being a diagnosis of the problem. in 1998. Because flooding is the result of runoff. Such flash floods present major risks to life. moves across the catchment. In areas subject to high levels of soil erosion. this they also deposit. However. it is necessary to consider controlling erosion and the patterns of sediment transport and deposition at the same time as flood mitigation. Two obvious approaches to flood mitigation are therefore to limit runoff and slow drainage. conclusions. Thus. The views.
Peaks increased by 7% in North Carolina (Hewlett and Helvey 1970). the combination of climate.World Commission on Dams Assessment of Flood Control and Management Options 20 The consequences of afforestation and deforestation Afforestation normally results in increased evapotranspirational losses. This reduction in flood protection may be offset by degradation of ditches dug during planting. Ditches increase drainage density and can halve time to peak. Finally. but only 50% in small basins. and higher annual runoff (Robinson 1990). However. produce peakier response. Experimental evidence from the Institute of Hydrology Plynlimon catchments shows 14% lower annual runoff in the forested basin (Blackie and Newson. 1986). Hofer (1998) suggests this is not such an important factor in increased flooding in Bangladesh. basin characteristics. e. -2%/10% increase in forest cover (Nisbet.1990). Climatic effects include: 1) less impact where torrential rains are common. higher flood frequencies and volumes. Removal of pine/eucalyptus forest can increase runoff by 40mm/10% change in forest cover compared with 10mm/10% for scrub (Dunne and Leopold 1978). tree species and husbandry cause wide differences in response. Jones and Grant (1996) found a 100% increase in peak discharges in large basins. Sometimes flow can be reduced by rapid grass growth. Increased sediment yields may aggravate flooding downstream. it may increase it if drainage ditches are used. but in basins where saturation overland flow is a dominant source of floodwaters effects may be less. conclusions. Afforestation may increase evapotranspiration by 50-100% and trees with high evapotranspiration rates may be planted specifically to reduce flood risk and lower annual runoff. or lower growth rates (Hudson and Gilman 1993). This will enhance infiltration-excess overland flow. and 3) reduced capture of ‘occult precipitation’ where forests were above the condensation level. but it does not necessarily reduce flood risk. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . others only on some peak flows (Wright et al. Some catchments show no effect. University of Aberystwyth This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. However. forestry practices can be extremely important and effects can be reduced by careful management (Collins and Pess 1997. especially root density. perhaps because of reduced interception rates through loss of branches or acid rain damage. Forestry Commission 1993).g. landslides. Deforestation tends to reduce evapotranspiration and runoff concentration times and thus increase annual runoff. which is commonly blamed on deforestation in the Himalayas. the effects may only have a ‘half-life’ of 2-7 years in high rainfall areas (Hibbert 1967). 13% (in winter) in British Columbia (Henderson and Golding 1987) and 38% in Malaysia (Bruijnzeel 1990). indeed. The views. The effect seems to reduce as the forest matures. or through changes in soil processes. 2) reduced snowmelt flooding where snow accumulation under the forest has been significant. 1990). perhaps because very intense storms in saturated basins generate substantial flows whatever the landcover (Bruijnzeel and Bremmer 1989). However. Infiltration capacities tend to reduce. and larger peaks may not be affected. J A A Jones. flood peaks and the peakedness of stormflow (Trimble and Weirich 1987. Bosch and Hewlett 1982). Flooding may be enhanced by accelerated soil erosion. through compaction by logging vehicles and new roads. Changes in evapotranspirational losses vary with species of tree. if natural regeneration is allowed. deposition and reduced channel capacities (Chan and Parker 1996).
Once other planning constraints are taken into account. The views. From the catchment perspective. Green 1999).World Commission on Dams Assessment of Flood Control and Management Options 21 2. few are untouched by human activity. a rise in average flood losses can be quite consistent with an increase in economic efficiency (Green et al 1993. water was plentiful. From an economic perspective.000 per property (Green and Warner 1999). the cost of providing infrastructure to a greenfield site amounted to some £10. Consequently. the floodplain may be the least damaging place for intensified development. The Wentlooge Levels are a coastal polder in South Wales. Closer examination revealed that the reason why large new investments are being attracted to the area are exactly the same as the reasons why the area was first reclaimed in Romano-British times and was a centre for industrial activity in the C19th: the only alternative land is either steep hillside or narrow river floodplain (Chatterton et al 1993). of archaeological significance or ‘Green Belt’. the municipal government decided to undertake flood alleviation works on the Black Brook near Loughborough and then to develop the area because the alternative was intrusion into Charnwood Forest (Parker 1995). From this catchment perspective. By contrast. and • arable land per capita. what is an appropriate policy in one country may be quite inappropriate in another. alluvial deposits. Three measures of the pressures and constraints under which an appropriate flood management policy must be developed are: • Gross Domestic Product (GDP)/km2. Indeed. rivers were the best routes for transport. the cost of renovating the existing dike system was around £4. such as designations as areas of landscape value. the land is flat. The constraints on flood management and the use of floodplains vary significantly between different countries. • population density. For example. the quality of life that could be achieved was greater. Thus. Whilst settling on the floodplains exposed them to the risk of flooding. that greater prosperity also reduced their vulnerability to other hazards. the term ‘floodplain encroachment’ is highly misleading and also carries an essentially ideological message: the real decision is whether it is better to develop on the floodplain than elsewhere. 3/4s of the population lived south of the Yangtze from rice farming. and the local wetlands provided good sources of material for building and domestic uses such as baskets. In relative terms. conclusions. Maddison (1998) reports that in C8th China. it looked like an ideal candidate for managed retreat.000 per property if the costs were only shared across households. This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. in the Wentlooge Levels. the costs of flood alleviation are often much lower than the infrastructure and other costs of intensifying development elsewhere. and recommendations contained in the working paper are not to be taken to represent the views of the Commission .2 The Use Of Floodplains Floodplains were amongst the areas first developed for human settlements: the soils are often rich. since the offshore mudflats are a Ramsar site. there are two different questions to be answered: • should further intensification in the development of the floodplain be encouraged? and • should flood alleviation be undertaken for that development? In most of the world. 3/4s of the population lived in north China from dryland farming. floodplains have major competitive advantages for human settlement and people from early history chose the advantages of arable farming on the floodplains over a poorer life hunting-gathering in the hills above the floods. The answer may be yes for one of two reasons: it is better to develop on the floodplain than anywhere else or there is nowhere else to develop. the floodplains are already part of the web of the socio-economic system. by the end of the C13th. For instance. This competitive advantage has continued to the present day. the objective is to make the best use of the catchment as a whole: it is not to minimise flood losses. This shift allowed an immediate doubling of the population together with a 30 per cent rise in per capita income.
and recommendations contained in the working paper are not to be taken to represent the views of the Commission .639 7 This is a working paper prepared for the World Commission on Dams as part of its information gathering activities.685. in global terms. a not particularly densely populated country. the second is a measure of the intensity of demand for those resources. Conversely.310. resettling people who live on the floodplain is a more practical option in those countries.387 331 The United Kingdom 3.World Commission on Dams Assessment of Flood Control and Management Options 22 The first is a joint measure of the availability of natural resources and the intensity with which they are already being used.552. and in food terms arable land is much more productive than grazing land.901 78 France 1.535 120 Montana 36. the lower are the values on the first two measures and the higher value on the third.095 2 Nepal 21.713 238 Germany 3.341 871 South Dakota 70.100 443 Japan 5. conclusions. and the third a measure of the availability of a key resource (Tables 7 and 8). The views. particularly those of C19th settlement.108 29 Bangladesh 152. the population of the USA would have to increase the equivalent of the 3/4’s of the population of China. Thus. In particular. where population densities are.859.760 2 Australia 33.930 143 Mali 1. it is less likely to be possible to consider abandoning some of that arable land or converting it to grazing. To bring the population density of the USA to that of France.186 28 Missouri 587. the less are the likely constraints in developing a flood management strategy.600.090 4 China 46.010.902 104 United States 596. where arable land is scarce. Similarly.702 228 The Netherlands 6.423.644.826 228 Illinois 1. very low. Table 7 Gross Domestic Product and population density per square kilometre Country or State GDP/km2 Population/km2 New Jersey 10.810 371 Connecticut 6. 100 million people inhabit the floodplain of the Yellow River in China (World Bank 1997).
World Commission on Dams Assessment of Flood Control and Management Options 23 Table 8 Arable land availability per capita Country Arable land (hectares) per capita Algeria 0. conclusions. the occupiers of the floodplains have adapted to the risk of flooding so as to cope with the flood hazard. as the song itself says.or mudslide. For example. with development. This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. They typically have the least choice where to live and end up by settling on the most marginal land. Drijver and Marchand 1985). and recommendations contained in the working paper are not to be taken to represent the views of the Commission . a similar wall has been constructed across the door between the living area and the bathroom so that when the toilet overflows in a flood. Thus.67 World average 0. Maltby 1986) are added to this equation. Nevertheless. At the same time.68 China 0. This is often on the steep slopes where. In some cases. Dixon et al 1994. lowland floodplains are frequently wetlands.24 source: World Bank Selected World Development Indicators 1999/2000 Precisely because they were settled early.17 Japan 0. At the same time.03 Thailand 0. they also provide fish and materials for construction and everyday use (Acreman and Hollis 1996. the living area is protected (Penning-Rowsell and Fordham 1994). The most extreme form of adaptation is perhaps that of the char dwellers of Bangladesh. ‘flood fighting’ and ‘flood defence’. others are inhabited by the least advantaged groups. These adaptations vary from the raised earth mounds constructed in Zeeland as flood refuge areas. These are amongst the richest habitats in the world and the strongest argument against intensifying development of the floodplains is usually the ecological value of the existing wetlands. Maltby 1986). The language of the consultants was of ‘flood control’.10 Germany 0. defining floods as something to be fought against and subdued as opposed to the celebratory and coping terms used in the local song. The views. Here. it can be more efficient to leave the wetlands alone. some floods are so extreme that they overwhelm the traditional coping responses. the slopes become even more unstable and there is a severe risk of a land. The value of wetlands in terms of providing fisheries and other functional values have now been extensively reported (Barbier et al 1997.10 USA 0. In Setubal.27 Australia 2. creating and eroding islands and the char dwellers retain title to land whether or not it is currently part of the river channel (Schmuck-Widmann 1996). floodplains are typically integrated into existing agricultural and economic activities. Crowards and Turner 1997. When the functional values of wetlands (de Groot 1987. the rivers are constantly changing their courses. many populations are highly adapted to the routine pattern of flooding. In other cases. Whilst some floodplains may have been chosen as highly desirable places to live. Floodplains are used for flood recession farming and seasonal livestock grazing. residents have adapted to frequent flooding by closing off their front door with either a steel door or a concrete wall.14 India 0.29 United Kingdom 0. in Portugal. Charles Namafe (1989) contrasts the language of a traditional song to the militaristic language of Dutch consultants studying a proposed polder scheme. to raising housing on stilts in Malaysia to the practice of taking refuge in roof areas in Bangladesh.
particularly where proximity to a river has no advantages to offset the risks of flooding. Floods are a typical Lozi’s patelo the royal drums are never bought they cannot be priced they cannot be given away. conclusions. The views. Floods are ours the floods themselves they know their own route they know their home area they know where they’re needed And when we ourselves see them we are inflated with happiness our hearts become lighter we do not fear floods. Floods in Bulozi (Western Zambia) translated by Charles Namafe from Sibetta O K (1983) Fa Munanga Wa Lyambai. they are drunken with cold air they watch a scene which comes but once a year floods are beautiful. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . Lusaka: Neczam It is floodtime in Bulozi! There is the floodplain clothed in the water garment Everywhere there is water! there is brightness! there are some sparkles! waves marry with the sun’s glory Birds fly over the floods slowly. Floods are a typical Lozi’s patelo When floods are in.World Commission on Dams Assessment of Flood Control and Management Options 24 the marginal land may be on the floodplain. We might give away cobs of maize or fishes This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. we prepare the royal boat It is a happy occasion in Bulozi Listen! the royal drums boom in the palace the royal drums are calling paddlers they are calling the youth paddlers and others. Bulozi is the floods’ place of abode every year the floods pay us a visit A Lozi does not beg for floods We do not turn to herbs to have floods We do not practice witchcraft whatsoever They are floodwaters indeed! The floods know their home area.
Floods. The debris from the shacks. Kuomboka Ceremony and the royal drums are all ours.World Commission on Dams Assessment of Flood Control and Management Options but royal drums are ours. In the semiarid regions a dry river bed can become a raging torrent within a matter of hours.000 residents living within a four square kilometre area. 25 Disadvantaged communities and their location in floodplains – a case study The unplanned and unmanageable large-scale migration from rural to urban areas is a feature of many developing countries of the world. may also block bridge openings and deflect the flood to an adjacent area that would otherwise have been out of danger. Lives may be lost when spectators gather on bridges or on the river banks and their escape routes are cut off as the river rises. It will be difficult and dangerous to use rubber boats on the river to rescue people trapped in the debris. Most local authorities in South Africa prohibit residential occupation below the 1:50-year floodline. and undermine the river banks causing the shacks on the banks to collapse into the river. This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. This will also hinder rescue attempts and increase the probability that people isolated in their shacks or washed into the river will drown. and ours alone their boom sound tickles our blood we get mad of our cultural heritage we then dress in animal skin loins. Flood levels and water velocities are very much higher than those in otherwise equivalent rivers in moderate climates. In some cases shacks have been built on refuse dumps within the channel itself. Alexandra has some 350. Often. the only available areas close to employment opportunities. The views. The Jukskei River through Alexandra has a steep gradient. and ours alone. A major flood will rise rapidly. thereby preventing helicopter rescues. The high velocity and large quantities of floating debris will prevent boat rescues. Floating debris. or the river banks collapse. particularly floating timber and submerged corrugated iron sheets caught in the fast flowing water.000 shacks are located within the stormwater drainage system inside Alexandra. There will also be a very short time for mobilising the rescue services. A hydrological factor that exacerbates the problem in the semiarid regions of Africa is that the variability of river flow increases with aridity.000 people living along the banks of the Jukskei River in flimsy shacks below this floodline. particularly large trees from the upstream catchment. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . The steadily increasing occupation of these flood prone areas by socially and economically disadvantaged communities in developing countries. so the water velocity during floods will be high and will cause severe damage to all structures in its path. It may still be raining heavily when floods occur. Shacks have been built on all available space right up to the edge of the almost vertical river banks. schools and hospitals. will seriously injure escapees attempting to wade through the water even if this is less than knee-deep. yet in Alexandra there are more than 6. destroy all shacks in its path and result in a large loss of life. Even a minor flood that does not overtop the river banks will engulf the shacks within the river channel. Another 3. conclusions. especially in Africa. A good example is the township of Alexandra within the Greater Johannesburg Metropolitan Area in South Africa. has resulted in an increasing proportion of their populations becoming vulnerable to floods. are on the floodplains that were previously identified as being unsuitable for residential occupation. Once the flood water level rises above the river banks the densely packed shacks further from the river will start collapsing.
The unplanned occupation of flood prone urban areas by socially and economically disadvantaged communities is a problem that is more likely to increase than decrease in the years ahead.so it is imperative that the communities at risk should know what to do when the sirens are sounded. Areas that have been evacuated may subsequently be re-occupied if the local authority does not have powers to prevent this happening. University of Pretoria. J. If they are given high priority. there are difficult political and economic decisions that have to be taken before this objective can be achieved. W. The Johannesburg Civil Protection Service and the Department of Civil Engineering of the University of Pretoria developed a simple flood warning system for Alexandra. R. This can be achieved by the provision of new houses in safe areas for those most at risk. It must be appreciated that flood risks are not the only risks to life and property that these communities have to face. Sirens within Alexandra will be activated by radio from the crisis centre when the water level in the river at an upstream road bridge reaches a level that is likely to pose a threat in Alexandra.possibly less than 30 minutes .World Commission on Dams Assessment of Flood Control and Management Options 26 This leaves the implementation of a flood warning system as the only viable short term solution. Those most at risk are usually those who arrived last by which time no other land was available. The only viable long term solution is to provide incentives that will encourage the threatened communities to move to less vulnerable areas. and other facilities. The advance warning will be very short . and member of the United Nations Scientific and Technical Committee on Natural Disasters. conclusions. South Africa. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . They therefore have the lowest priority for new houses. schools. rather than to move to better housing further from their places of employment. Alexander. This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. The views. However. This has the following components Automatic rainfall telemetry equipment was installed at several sites in the Jukskei River catchment. Forced removal is politically unacceptable and occupants of shacks in unsafe areas may prefer to stay where they are. this policy will become known and will encourage others to deliberately occupy unsafe areas. This information is relayed to the crisis control centre. Department of Civil Engineering.
Thus. at one time all flood impacts were ‘intangibles’. It has been found quite consistently This is a working paper prepared for the World Commission on Dams as part of its information gathering activities.e. lead to medium to long term gains in industrial efficiency where plants based upon out-dated layout and machinery or replaced by redesigned and updated ones. Tangible flood losses can be very large as in the 1988 river flood in Bangladesh which caused an estimated US$1. and lead to an increase in family and community spirit and bonding. some have been converted into ‘tangibles’. The Benefits and Costs of Floods The effects of floods are extraordinarily complex and include both beneficial (i. regions heavily affected by floods may lose trade to neighbouring regions which remain flood-free.al.World Commission on Dams Assessment of Flood Control and Management Options 27 3. negative) impacts on society and the environment. although such measurement is hardly ever precise and relies heavily upon damage estimation procedures. replenish soil moisture which may be a factor in subsequent increased crop yields (this is particularly important in ‘recession’ agriculture). as the state of the art has developed.e. One common way is to categorise flood losses as tangible or intangible. Parker and Thompson 1991). One of the problems of flood impact assessments is that they often focus solely or mainly upon the adverse impacts and a truly balanced flood impact assessment methodology has yet to emerge. For example. Some of the most beneficial social and environmental effects of floods are listed in Table 9. Following damaging floods there is usually an increase in business amongst building and repair firms who gain from the flooding. and the increases in economic and other activities which are generated by floods. These are effects which may be hidden and forgotten. Similarly. be beneficial to aquatic ecosystems and to human livelihoods (e.g. they may be the most or less important impacts of flooding. the loss of an archaeological complex by erosion caused by flooding). positive) and adverse (i.416 million (Islam 1997). They arise from the moisture and sediment content of floodwater. Tangible damages are usually taken as those which can be measured in monetary terms. The views. and/or those for which monetary estimates are considered undesirable and inappropriate. such as the dollar damage to the fabric of a factory. There is no doubt that floods destroy economic resources of value to society. Flood impacts are ‘distributive’ in that losses in the floodplain sometimes generate gains elsewhere outside of the floodplain. 1987. conclusions. Table 9The beneficial impacts of floods Among the beneficial effects of floods are that they may: • • • • • replenish soils with alluvial silt which adds to soil fertility and subsequent soil productivity. At any time.g. but much is now known about these impacts (Parker et. but it is also true that floods can generate financial gains which can offset financial losses (Parker 2000). the flooding of a retail outlet will normally lead to customers temporarily transferring their business to a flood free outlet which experiences a gain in business. therefore. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . the positive impacts of floods on aquatic systems. fishing) associated with them. ‘intangibles’ are the bits left out of the analysis and because they are not quantified. The adverse effects of floods are difficult to trace because of the multiple-order effects on society and economy. Intangible losses are those which either presently defy monetary measurement (e. The impacts of floods may be divided in several ways.
Direct flood losses are those which arise through the physical contact of floodwater with people or property. Figure 3 Subjective assessments by households of the relative severity of the different impacts of flooding (sources: FHRC data) relative severity of flood impacts 10 9 8 7 severity 6 5 4 3 2 1 0 Disruption and recovery Stress of the flood Worry about future flooding Having to leave home Loss of memorabilia Damage to house Damage to replaceable items Effects on health area flooded scales: 0 = 10 = minor impact very severe impact A further way of dividing flood losses is to categorise them as direct or indirect. when a factory suffers direct flood damage its production processes may be interrupted leading to loss of sales and other losses (such as increased cleaning costs). and recommendations contained in the working paper are not to be taken to represent the views of the Commission . secondary and tertiary losses are identified. disruption and loss of items of sentimental value . The views.the stress. Penning-Rowsell et al 1992) that the ‘intangible’ impacts of flooding on households . as for example in floodwater damaging the carpets and furniture of a home through immersion. This is a working paper prepared for the World Commission on Dams as part of its information gathering activities.are more important to the affected households than the damage to their home and its replaceable contents. 1989.World Commission on Dams Assessment of Flood Control and Management Options 28 (Allee et al 1980. The results from more than 1300 interviews with flood victims are summarised in Figure 3. conclusions. For example. a third way of categorising losses is through multiple-order effects so that primary. Finally. Green and Penning-Rowsell 1986. Indirect flood losses are those damages which are consequent upon direct flood damage. The impacts are arranged from front to back in terms of increasing average severity and the locations similarly ordered from left to right by decreasing average impact severity.
with damage values tending to increase as velocities increase. conclusions. Oi (1993) shows a 5.World Commission on Dams Assessment of Flood Control and Management Options Figure 4 The mechanisms of flood losses 29 depth sediment salt chemicals duration flood losses loads sewage debris depth + velocity battering by debris velociity undermining of foundations Flood damage is produced by a number of ‘damage mechanisms’ (Figure 4). Whilst the interaction of depth and velocity is important in determining the probability of structural collapse. In residential areas floodwater is usually contaminated with sewage because of the flooding of sewers and sewerage systems. damaged structures and vehicles caught in the flooding) which are themselves mechanisms of further damage. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . New South Wales Government 1986. Black and Evans 1999) and depth-damage curves have been developed for a number of countries (Figure 5). they can create locally intense flood velocities that induce scour that can itself induce the structural failure of buildings (US Army Corps of Engineers 1998a). for example. This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. Riverine flash floods and dam-break floods can generate high floodwater velocities which are quite capable of sweeping away robustly constructed structures and buildings (Bureau of Reclamation 1988. Sangrey 1975). The depth of flooding is a key determinant of flood damage (Penning-Rowsell and Chatterton 1977. This can lead to fire or. timber and masonry buildings are likely to fail when flood velocities exceed 2 metres/second. The sediment or debris load of floodwaters is one example and floodwater is also often contaminated. Floodwater velocity is also a determinant of flood damage. In industrial areas floodwater may be contaminated by any manner of chemicals derived from commercial processes and vehicles. both direct and particularly indirect losses. Flash floods often carry large loads of debris (derived from trees. The views. to the contamination of fresh water supplies. Indirect industrial flood losses are related to the length of plant outage. A number of other mechanisms may contribute to overall damage values. Flood duration can lead to increased flood damage values. Emergency Management Australia 1999. For example. eroded banks. both as a result of closure whilst flooded and recovery afterwards (Parker et al 1987).000 tonne boulder that was deposited by the Mandu Khola during the 1993 floods in Nepal. Reiter 2000. Since buildings obstruct flood flows.
1 Risks to Life and Health Floods are high probability events since protection is not usually provided to reduce the risk to less than 1 in 200 per year and in most countries the standard of protection is considerably lower.50 1. in 1976 and flash floods in Southeast China in 1996 (Gruntfest.00 0.2 1.World Commission on Dams Assessment of Flood Control and Management Options Figure 5 Depth-damage curves for dwellings (Source: Halcrow et al 1999) 30 Dwellings: structural losses 4.50 0.SMEC 1975 Canada.00 0. the risk of death should a flood occur is usually low compared to other hazards.IBI/ECOS 84m2 2storey* (1982 prices) depth 3. and where very large areas of low lying land are affected.3 UK:FLAIR .4 2. October 12.00 3.75 2. The views.74 3. Very often in major events the number of lives lost may never be accurately known.00 proportional loss 2.8 0.50 3. However. 1999)). large numbers of people may die. Nevertheless.all dwelling USA . conclusions. and recommendations contained in the working paper are not to be taken to represent the views of the Commission .1 0.50 4.FIA 1985 (Appelbaum) Japan % Bangladesh : house type 4 Australia . floods do kill and sometimes kill hundreds of people (floods in Eastern Mexico in October 1999 resulted in at least 300 deaths (Independent. USA.00 1.50 2.3 1. 1997). with very large populations. • flood onset is sudden as in flash floods. This is a working paper prepared for the World Commission on Dams as part of its information gathering activities.66 -0. The flood conditions in which the risks of death are likely to be greatest are those where one or more of the following conditions exist: • flow velocities are high. for example the Big Thompson flood. although this depends on the zone affected. as in the Bangladesh floods of 1987 and 1988.
therefore. ground floor apartments. Timber framed buildings. into excavations or into ditches. people attempting to move about. is increasing this risk. especially from flash floods but also from burst water mains and surcharged sewers. For example. to be consistent with this threshold value. the conditional risk of death should a flood occur cannot exceed one in ten thousand: there is no doubt that in some contexts the conditional risk of death is considerably greater than this. conclusions. or migrant. Thus. as in Bangladesh. the risk to life from flooding is likely to be higher than those levels of risk which are deemed to be acceptable or tolerable in regard to such hazards as nuclear power stations or chemical plants. Reiter 2000) or stuck in the flood water. as does Finnish research (Reiter 2000). Deaths can also occur where people are trapped in single story buildings. concluded that the safe limit would be a product of depth (metres) times velocity (metres/sec) of 1. collapsing or being swept away (Green. Emergency Management Australia 1998. and where extensive low lying densely populated areas are affected. Many deaths in floods occur because people attempt to drive though or away from flood waters and get swept away or trapped in their cars. Flood alleviation and other artificial structures themselves involve a risk to life because of the possibility of failure. The growing tendency to multi-levelled cities where shopping centres and cinemas are below ground level. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . Eighty percent of the estimated 200 deaths in Monterrey. temporary and fragile structures and tented dwellings may give rise to significant loss of life or hazardous rescues. Parker and Emery 1983). Abt et al. It is. essential to assess whether the risk of death is particularly high in any area under study and to determine what are the most appropriate measures to This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. (1989). In flooded urban areas. natural or artificial protective structures fail by overtopping or collapse.0. their cars either then get swept away as a result of positive buoyance (Bureau of Reclamation 1988. such as railways or car parks which can pose a particular threat to life in urban areas. The views. New South Wales Government 1986. In many countries. informal. In Bangladesh. Metro systems present a particularly high risk.World Commission on Dams Assessment of Flood Control and Management Options • • • 31 flood waters are deep. Pedestrians unaware of the power of flood waters may be swept away. mobile homes. it is commonly very high relative to that from other hazards. it is unusual for a flood alleviation project to be designed to protect against a flood more severe with than that with a return period of 200 years. Consequently. a general rule of thumb has been adopted that an individual risk of death per year of one in one million is a threshold value. New South Wales Government 1986). people who are likely to lack the resources to build sound structures. Australian data give similar results (Emergency Management Australia 1998. floodplain land is the only space available for settlement particularly by poor. Outside of the Netherlands and some other countries. For those other hazards. Mexico in 1988 were attributed to attempts to ford the flooded river (Vazquez et al 1997). Deaths occur in flood events through a range of circumstances: • Death rates in floods are high where buildings fail to provide a safe refuge (Table 10). may fall down blown manholes. for example dam or dike failure. • • • • • • It should be stressed is that where there is a risk of flooding. in an experimental study. People trapped in buildings or on the roofs of buildings may die from exposure as illustrated by the Mozambique floods of February/March 2000. cellars or underground structures. a significant number of deaths during floods are from snake bite as both people and reptiles take refuge in the same trees. in the Big Thompson flood in USA many of those who died were drivers who attempted to outrun the flash flood. as for example in Hong Kong. particularly where flood waters are turbid or discoloured.
1983. Powell and Penick. and respiratory illness due to exposure to poor and insanitary living conditions. higher than expected cancer rates. Research literature on floods (Bennet. Handmer and Smith 1983. Thus. events may pose a serious threat of illness and death through gastro-intestinal illnesses. 1994). This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. 1985. although the evidence on this is inconclusive. 1970. flood warnings. However. 1983) indicates that these events can have significant health effects. 1994) although the evidence is mixed as to the differential vulnerability of the elderly. Penning-Rowsell et al 1992. 1986. Blaikie et al. 1994). the old. Green et al.World Commission on Dams Assessment of Flood Control and Management Options 32 reduce it. a major cause of subsequent health damage seems to be the stress of the flood itself. and the disruption to transport. the contamination of drinking water supplies. Recent qualitative research (Tapsell et al. There may be an increase in the incidence of malaria. of flood victims. Koopman and Spiegel 1998) and on dam bursts (Baum. 1999) suggests that the stress of the flood event itself. Heurta and Horton. conclusions. Waelde. 1999).. the sick. Tapsell et al 1999.0002 for low severity floods in areas where there is a good understanding of floods and more than sixty minutes of warning to 0. yellow fever and other insect and snail borne diseases as these breed in standing water (Blaikie et al. cholera. as well as the well being. food and medical supplies and services (Gueri et al. Some analyses have been undertaken of past floods (DeKay and McClelland 1993. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . dysentery. the stress and disruption to life during the recovery period and worry about future flooding can have a serious effect on physical and psychological health. Drescher and Abueg 1995. the disabled and the poor (Blaikie et al. Graham’s (1999) analysis gives conditional probabilities of death that range from 0. The pervasiveness of morbidity and disablement problems after flooding is indicated by Sikander’s (1983) finding that between 43% and 57% or rural households surveyed in Pakistan following flood events fell ill after floods. the provision of flood refuges or evacuation plans. Depending on the circumstances and characteristics of the flood. illnesses and injuries may result in raised death rates in the aftermath of flooding (Gueri et al. ranging from premature death.75 in areas just downstream of a catastrophic dam failure occurring without any effective warning. 1986). The views. sewage spread. These measures may include one or all of the following: emergency plans. There is evidence that stress induces immunological changes. Those most vulnerable to health effects of flooding are likely to be those who are most vulnerable in normal conditions: the young. Kraak 1994) in order to try to estimate the risk of life from flooding but it is difficult in such studies to separate out the small part of the population who were exposed to a high risk from those who were exposed to a much lower risk. Pre-existing medical conditions were perceived to be exacerbated and some new mental and physical conditions were attributed to the flood event (Tapsell et al. 1978. and other clinical problems requiring hospitalisation and medical consultations..
The views. the appropriate response is not to undertake a flood alleviation scheme but. one such form of irrigation is then recession planting (Marchand 1987). and on average flooding. although floods may also deposit sand that decreases soil fertility. this is of little relevance to the subsistence farmer who cannot fed their family in the coming year because the crop has been destroyed as the result of an extreme flood. In temperate regions. conclusions.World Commission on Dams Assessment of Flood Control and Management Options Table 10 33 Flood conditions and probability of structural failure of masonry buildings (Source: Penning-Rowsell et al 1992) Result Velocity Depth x velocity < 3. to introduce systems of low cost credit and other ways of reducing the risks of crop loss. In these circumstances the soil becomes saturated and this gradually destroys crops. Damage occurs when the water table rises and begins to damage crop roots. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . localised flooding from local rainfall may cause flooding and crops losses in an area before that area has This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. Rogers (Rogers et al 1990) argued that Bangladesh did not have a flooding problem but an irrigation problem.2 Agricultural Benefits and Costs Clearly flooding can have highly positive effects in agricultural areas. the activity of blue-green nitrogen fixing algae during the floods are considered to play an important role in increasing yields from the post-flood crop (Brammer 1990). semi-arid. Thus. having borrowed against the expected crop to buy seed and fertiliser.0 metres2/ second /second <7. both too much and too little water being harmful to plant growth and yields. and the ‘normal’ annual flooding which takes place in many agricultural floodplains of the world is a vital source of moisture for crop growth. may be beneficial. In addition.0 metres2/ inundation damages only <2 metres/second partial failure > 2 metres/second structural collapse > 2 metres/second >7. Alluvial silt from flooding is also generally a rich source of additional fertility (Table 9). it is perhaps better to think of agricultural areas in terms of water level management rather than of flooding. This is particularly so in arid. frequently. arguably in these circumstances. whilst in the long run. an excess of soil moisture is a barrier to agricultural productivity and the management problem becomes one of maintaining the water table at an appropriate height relative to the root zone. In Bangladesh the agricultural economy depends upon regular flooding which moisturises soils. In Bangladesh.0 metres2/ second > 3. Nor is it to the farmer who. tropical and sub-tropical climates where drought is a major problem and irrigation is a regular requirement for crop growth.0 metres2/second 3. In general. But. the larger part of the value of the crops has been lost before ‘flooding’ occurs. However. looses their land and is reduced to a agricultural labourer or share cropper. for example.
However. the maintenance of efficient drainage networks and the installation of under-soil drainage. While. Thus. local flooding within the protected area can destroy the crops notionally protected by the dike. crop losses are typically much lower than those to be expected in urban areas and it is correspondingly more difficult to justify providing flood mitigation in rural areas or to provide a high standard of protection. The views. In areas where there are two or more crops each year. Therefore. conclusions. irrigation may be needed in temperate climates during the drier seasons. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . There are often also indirect subsidies such as subsidised prices for electricity. simply looking at flood mitigation for agricultural land in terms of crop losses can be to misunderstand the reasons why the works were undertaken in the first place. fertilisers. In some cases it may be possible for the farmer to plant an alternative second crop but this will generally give a lower valued yield than the preferred crop.World Commission on Dams Assessment of Flood Control and Management Options 34 flooded from the major river in the area. agricultural production is also typically directly or indirectly subsidised.2. 3. Direct subsidies include guaranteed prices. They may be driven by concerns about alleviating rural poverty. From an economic perspective. In terms of loss per hectare.1 Flood Losses One of the difficulties about policies relating to agricultural areas is that they are only rarely primarily about agricultural production. This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. The extent of the crop losses to be expected depend upon where the flood occurs in the growing season and also on the duration of flooding (Figures 6 and 7). For agriculture. or simply securing votes. but they are not generally about making the most efficient use of agricultural land. the value of reducing the amount of crops lost through flooding can then be rather low since increasing crop production will also increase the amounts that must be paid in subsidies. the distinction between land drainage and irrigation is somewhat artificial and it is better to think in terms of water level management: of managing soil moisture levels within the optimal range. For these reasons. in areas protected by a dike. If soil water levels are not controlled then there is generally little advantage in controlling surface water levels. a flood may damage or destroy not only the standing crop but also prevent or delay the planting of the next crop. in many cases.g. fuel and irrigation: it being rare that farmers pay all of the costs of irrigation (Garrido 1999). the crops that destroyed by a flood had already been severely damaged as a result of water-logging before the land was eventually flooded (e. as found in the Common Agricultural Policy in Europe. Hong and Zhang-Yu 1999). land improvements. avoiding migration to the cities. ensuring national food security. Land drainage comprises a range of techniques designed to keep water levels below the root zone such as continuous pumping of land. in the wet season land drainage will be required to lower the water table to acceptable levels for farming.
2. This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. which only provide protection against the floods in the peak growing season have commonly been provided ahead of the main dike line. the rationale for land drainage has been to allow a switch from rough pasture to arable uses. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . through credit or other means. The views. between good and bad years.World Commission on Dams Assessment of Flood Control and Management Options Figure 6 Crop losses in a northern climate (Hungary) (Source: Podmaniczky 1999) % loss of yield . provided that they have a way of bridging.2 Land Drainage In agricultural areas it is typically land drainage rather than flood alleviation that offers the major benefits.flood duration of 15 days 100 90 80 70 60 % 50 40 30 20 10 0 March Sept May Jan Nov July 35 Cereals Fodder crops Maize Sugar beet Sunflow er Grassland m onth Figure 7 Crop losses in a Southern climate (Australia) (Source: Higgin 1981) losses as % of gross margins 100 80 % 60 40 20 0 WHEAT SORGHUM BARLEY/OATS SOYABEANS MAIZE SUNFLOWER month RAPE SEED 3. ‘summer dikes’. Controlling soil moisture levels increases yields and allows farmers to switch to higher valued but more sensitive crops: the consequent gains can be substantial (Table 11). conclusions. in Britain. in the Netherlands and Germany. Farmers will make this switch even though there continues to be a high risk of flooding. agricultural areas were seldom protected against floods with a return period of more than 10 years. Thus. Similarly. Historically.
Such problems have not been uncommon in Bangladesh. they may impede the drainage of the protected area. is that a land drainage scheme can have negative distributional consequences: small farmers may be unable to finance the necessary works and may be forced to sell out their land to the large landowners. In such systems. Land drainage also requires considerable on-farm investment in the form of field drainage. Furthermore.3m 3. Conversely.1997/98 prices) very bad -103 131 115 109 165 109 109 :change to LUT4 1 2 3 4 5 6 7 land use type extensive grass intensive grass grass/arable rotation all cereal rotation cereal/oil seed rotation cereal/ root crop rotation horticulture definitions of drainage good bad very bad water table depth from surface > 0. introduced into paddy fields. the herds of livestock are grazed on different areas at different times of the year. Fish and shellfish may be caught in the natural water courses.5 water table depth from surface <0. The views. and recommendations contained in the working paper are not to be taken to represent the views of the Commission .2. it will be those farmers who have access to capital who undertake the works first. conclusions. a large number of animals may be killed. aquaculture plays an important role. in extreme floods.World Commission on Dams Assessment of Flood Control and Management Options 36 Indeed. Indeed. Table 11 Economic benefits of land drainage in the UK (Source: Dunderdale 1998) drainage status good bad -73 -81 320 245 283 215 280 217 329 263 280 217 1500 750 economic net return (£/ha .5m water table depth from surface 0. Instead it will be some years before all farmers make that investment. One consequence. agricultural areas are commonly used for emergency flood storage. The dependency of arable production on soil moisture levels means that some agricultural flood alleviation schemes have no. the loss of a draught animal by a small subsistence farmer will have a serious effect on that farmer’s ability to plant and harvest the next crop. Both in traditional agricultural systems and in new systems. one consequence is that the conversion of the floodplain to arable land can seriously impact on way of life of livestock farmers. it cannot be expected that this will be undertaken by all farmers immediately that the land drainage scheme is completed. promoting conflict between the arable and livestock farmers. Moreover.3 Livestock Benefits and Losses In many traditional farming systems.3 to 0. the floodplains are important areas of rough grazing for part of the year (Marchand 1987). or negative. or raised in This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. benefits because they are not accompanied by a land drainage scheme for the protected area. therefore. both the likelihood that animals will die and the number of animals who will be killed increases with industrial farming methods.
Impacts will be greater for rapid-onset (flash) floods than for slow onset floods which may take days to develop and which allow time for rescheduling and contingency plans to be put into place. Effects on urban areas are likely to be greater with region-wide or city–wide flooding as opposed to flooding in one part of the region or city. Fish in particular may contribute a significant proportion of protein in local diets as well as being a major source of cash income. in countries with limited alternative sources). Flood losses arise from the complex inter-dependencies which exist in urban economies in which there will be a certain degree of redundancy (e. possible loss of exports Physical damage to retail and related businesses (e. leisure services. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . loss of production. The impacts of floods on urban areas are generally negative. may lead to food and other shortages in urban hinterlands (e.World Commission on Dams Assessment of Flood Control and Management Options 37 fish ponds. The views. conclusions.e. offices) Physical damage to distribution depots and systems Physical damage to manufacturing plant and equipment This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. transfers) Stimulus to rethinking the design and layout of plant or urban areas to reduce flood exposure and impacts (in extreme cases urban relocation from flood prone to flood-free sites) Increases in industrial and commercial efficiency arising from severe flood damage and rebuilding to more efficient standards Adverse (i. Disruption to manufacturing.g.g. duplicated or alternative systems) which will allow urban economies to overcome some of the effects of flooding. re-routing of incoming tourists to alternative flood-free destination Interruption of distribution systems. loss of sales. Table 12 Beneficial • • • • A categorisation of urban impacts of floods Financial gains in the manufacturing and service sectors arising from flood damages and losses (for repair and replacement of damaged goods) Financial gains in flood-free commercial sector arising from flood losses in flooded commercial sector (i. although transfers are bound to take place between flooded and flood-free sub-regional economies within metropolitan areas as a consequence of flooding. 3.3 Urban impacts Some of the more important urban impacts of floods are categorised in Table 12.g. flood losses) Direct Physical damage to dwellings Indirect Disruption of household activities Increased heating costs Costs associated with evacuation Disruption to trade and commerce.e.
fire. The exposure of urban areas to flooding is increasing rapidly over time. or both. The views. a flooded road artery is likely to generate considerable congestion throughout a large part of a city. This is because catchment urbanisation leads to the replacement of natural pervious surfaces with impermeable ones. (1997) Disruption to travel. Shanghai. Bangkok) lowering land levels. The growth of urban areas is also sometimes associated with increased convectional storm activity and increased convectional precipitation. medical services. nodes and links so that the overall vulnerability of the socio-economic This is a working paper prepared for the World Commission on Dams as part of its information gathering activities.e. conclusions. reducing soil infiltration and increasing runoff volumes. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . and well beyond the area physically affected by floodwater. The socio-economic system may be considered as a network made up of nodes and links (Green 1995). schools. in existing developed areas and newly-developing ones) are rising inexorably within cities. For example. and many of the world’s largest cities are developing in flood-prone coastal zones in which sea level is rising relative to land level. The susceptibility of urban systems to adverse flooding impacts is dynamic. At the same time urban drainage systems often operate in such a way as to reduce the times of concentration of runoff in streams and rivers. In many cities land subsidence is a problem (e. knock-on effects beyond the flooded area including possibly on health care Physical damage to public utilities Disruption of utility services leading sewerage and sewage treatment to utility ‘outages’ – knock-on effects gas distribution and storage facilities beyond the flooded area electricity generation and transmission water treatment.g. Similarly. ambulance and other emergency service installations Based on Parker et. At the same time the values of property at risk (i. Flooding of large cities can have impacts which spread out within the nation and globally. Flooding may then affect either. damaged electricity substations in a flooded area may well control the supply of electricity to areas beyond the flooded zone. hospitals.g. All of these effects can exacerbate the urban flood hazard and the impacts of flooding (Parker 1995). congestion and knock-on effects beyond the flooded area Additional costs of deploying the emergency services generated by a flood emergency The effects of floods in urban areas can be expected to spread well beyond the area immediately affected by floodwater because of spatial dependencies.) 38 Disruption of public services. and disruption of communications. storage and distribution telecommunication installations Loss of utility and/or physical damage to communication systems vehicles road systems rail systems waterway systems airports Costs of emergency services damage to police. especially if financial markets or tourist facilities are affected.World Commission on Dams Assessment of Flood Control and Management Options Physical damage to public buildings and structures (e. bridges etc.al.
vulnerability is decreasing: the switch to mobile phones is decreasing the vulnerability of telecommunications to flooding. the increased ownership.4. computers) increases susceptibility to direct flood damage but increases resilience to indirect flood loss. conclusions.World Commission on Dams Assessment of Flood Control and Management Options 39 system is a function of the number of nodes and links affected. the historical pattern of water level management has resulted in important ecological reserves (Purseglove 1988). thus contributing to the biodiversity of the river system as a whole. the susceptibility of the individual nodes and links to damage or disruption. Floodplain aquatic ecosystems systems depend on floods for their water supply. Floodplain rivers also have associated aquatic systems such as standing water and flowing water in branches which contain at least some species not found in the main channel. and dependence upon.4 Environmental consequences of flooding Ecosystems develop around the prevailing water regime. if one plant is affected it can be difficult if not impossible to source from alternative producers. it can be argued. it is necessary to differentiate between floodplains that are currently unregulated and those where previous flood mitigation strategies. For example: by covering grassland in sand. For example. The ‘Flood Pulse’ is argued to the major force responsible for maintaining the complex physical and ecological structure of a floodplain (SAST 1994). In some cases.1 Unregulated Rivers Floodplain rivers that are unregulated are flanked by alluvial floodplains which are subject to flooding during high flows and support biota adapted to floods. land drainage or irrigation projects have created a particular water regime. Floods. These systems are diverse and specialised and also support rare flora and fauna not found in the parent river. 3. or by isolating aquatic communities isolated when a river changes course after a flood. dissolved and suspended material is moved from the floodplain to the river channel supplying nutrients. Diverse and specialised plant communities are maintained in mosaics by the spatial and temporal patterns of flooding.g. During flooding. 3. The introduction of new technologies into urban life both reduce and increase susceptibility to flood loss – the picture is a particularly complex one. Therefore. and the structure of the network. High concentrations and diversity of terrestrial species are found in floodplain areas because of their plant productivity. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . Therefore. The views. However. washing fish fry away in floodwaters. floods are part of the natural This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. At the same time. water is delivered to the floodplain and as flood waters retreat. electronic appliances (e. Floods do cause specific damage as a result of ‘disturbance’ following flood events. The introduction of ‘just-in-time’ concepts into the planning and management of retail and other businesses means that these entities may be more susceptible than previously to flood loss (Parker and Tapsell 1995). Flooding stimulates the productivity of these systems and also allows for an exchange of organisms such as fish and resources between the river and associated wetlands. developed economies are characterised by lower degrees of redundancy than earlier economies: individual industrial plants both produce a larger proportion of total production and individual factories are more specialised. have a positive ecological impact by sustaining flood adapted species and supporting the ecological well being of the river ecosystem as a whole. In other respects. Mammals that live in the floodplain are usually amphibious spending time both feeding in the water whilst raising their young on land which may then be susceptible to flooding. organic material to the riverine food web. This appears to be the case throughout the world (Hillman 2000). use of.
World Commission on Dams Assessment of Flood Control and Management Options
hydrological cycle and in natural floodplains, the effects of a flood are short term (Haeuber and Michener 1998). Even some extreme events (for example, 1 in 100 year events) do not seem to cause more much drastic changes than lesser events (Hillman 2000). Extreme events are further argued to reset the dynamic functioning of rivers: whilst in the short term, some species are severely affected, others gain (SAST 1994). There is reported to be some published evidence on the ecological impact of a large dam collapse (SANDRP 2000) the high flows and long duration of flooding might be expected to cause damage and habitat loss down stream. Where the rivers that flood are highly polluted, for example the Rhine flooding of 1993, severe ecological damage can occur through contamination of floodplain areas by flood waters, sediment containing harmful chemicals and other substances (Dombrowsky and Ohlendiek 1998).
3.4.2 Managed and Regulated Rivers
Most floodplain rivers are managed to some degree for flood mitigation and flow regulation so that the interaction between the river and floodplain ecosystems may be interrupted or broken completely. The regulated release of flows from deep storage reservoirs may lower the temperature of rivers and alter the velocity of flows affecting riverine ecosystems (Acreman et al 1999; Bergkamp et al 2000). However, river management for flood mitigation or other purposes can have positive environmental impact. Established drained and protected areas may support rare species and special habitats. Flood retention ponds and reservoirs may also provide valuable habitats for example, for birds, and may make possible releases of water for environmental purposes, for example, to provide water for low flow rivers or drying wetlands. Because the ecosystems behind existing flood embankments and around existing river regulation projects have developed around the resultant water regime over, in some cases, hundreds of years, abandoning those projects can have very damaging environmental consequences. Thus, when the rehabilitation of the weirs on the River Nene was appraised, a large component of the benefits was contributed by the costs of otherwise maintaining the existing water regime in the large number of sites of local, regional and national environmental importance along the river (Balfour Maunsell 1995). Conflicts may further emerge between the proponents of the new ecosystem created by a new water regime and those who argue for a restoration of the previous water regime. Walters (1996) notes, for instance, that the maintenance of the new food chain in the Great Canyon resulting from the new water regime created by the dams may be threatened by the restoration of seasonal flooding targeted at the maintenance of the habitat requirements for endangered native fish species.
3.4.3 Offshore Impacts of Flooding
The sediment deposited by the freshwater plume of flood waters can have a deleterious impact on coastal species. Thus, flash floods in Sinai result in sediment deposition on the coastal coral reefs (Ras Gharib et al 1997); increased soil erosion and the loss of the coastal mangrove swamps causes the same problem in Barbados; whilst in Spain the channelisation of the rivers, and hence their greater capacity to carry sediment, has resulted in offshore ‘landslides’ and damage to sea grass beds.
3.4.4 Integrated catchment management and the environment
Holistic flood management is based on integrated catchment management across all functions and on the co-ordination of land and water planning. As the shift towards holistic management occurs, we have seen three waves of progressively bolder approaches to taking account of the environmental consequences of projects.
This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. The
views, conclusions, and recommendations contained in the working paper are not to be taken to represent the views of the Commission
World Commission on Dams Assessment of Flood Control and Management Options
1. Minimising the environmental harm done both by the works themselves and maintenance of those works (Purseglove 1988); 2. Incorporating environmental enhancements into the project (Wasserwirtshaft in Bayern 1990), restoration of some canalised rivers to a form nearer their natural one (Brookes 1990; Brookes and Shields 1996; Federal Interagency Stream Restoration Working Group 1998), and the use of bio-engineering techniques for bank protection (ARSIA/GEOPLAN/CEHIDRO/WASSER WIRTSCHAFT 1999); 3. The ‘Leitbilder’ concept (Bundesministerium fur Bildung, Wissenschaft, Forschung und Technology 1995): the development of comprehensive vision of the catchment as a whole in what would be its natural condition. For example, the ‘Stork’ project (de Bruin et al 1987) and landscape project in the Netherlands (Ministry of Transport, Public Works and Water Management 1996), and the “Loire Grandeur Nature” project in France (Gresent 1996)) The first two approaches are local in vision and limited to flood management whereas the third is a view of the catchment; the first is limited to minimising harm whereas the latter two are concerned with improving what is already there, where what is there has been degraded by past anthropogenic actions. The approaches are progressively more difficult to implement in institutional terms because no one institution has the authority to undertake or require all of the necessary actions.
This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. The
views, conclusions, and recommendations contained in the working paper are not to be taken to represent the views of the Commission
World Commission on Dams Assessment of Flood Control and Management Options
4. Flood Management Options
Flood hazard management is about managing risks under conditions of uncertainty. There is a risk of a flood of a particular magnitude and a risk that the management strategy will fail if that flood occurs. Dams may fail, dikes breach, culverts block or warnings fail to be given. Climate change is likely to change both rainfall patterns and intensities as well as runoff (Arnell et al 1994). Therefore, in planning a flood management strategy it is important to consider how floods will be managed when they are greater than the design standard of protection or the management strategy fails for other reasons (Green et al 1993). It is thus necessary to consider how they may fail; to apply reliability engineering approaches to the design of the management strategy. In managing the floods in one area, it is not helpful to do so by simply moving them on to elsewhere; to alleviate the flood problem in one area by making it worse somewhere else. The risks, and their consequences, are the subject of uncertainties. These outcome uncertainties are of two kinds: ‘parametric’ and ‘systemic’ uncertainty (Blockley 1980). The former has been called ‘what we know we don’t know’ and the latter ‘what we don’t know that we don’t know’ (PenningRowsell et al 1992). Parametric uncertainty covers, for example, the confidence intervals around the estimates from hydrological and hydraulic models as well the measurement error in the data used to calibrate those models. Systemic uncertainty then refers to the possibility that our theories and the models built on those theories are incomplete, false or inaccurate. Because choices are between alternative futures, and the future is unknown, all decisions are inherently uncertain. The meaning of uncertainty about outcomes is that it is not possible to completely differentiate between the outcomes; the ESRC (1999) drew the important distinction between uncertainty about probabilities and uncertainty about outcomes and they went on to categorise the different conditions under which decisions must then be made (Table 13). Table 13
Knowledge and probabilities
Knowledge about outcomes
Differentiated probabilities Undifferentiated probabilities
(Source: after ESRC 1999) Thus, when someone is thinking about playing roulette, then they are taking a decision under risk: both the outcomes and the probabilities of each outcome are known. When that person is thinking about which video to rent, they are making a choice under uncertainty. They know the possible outcomes, whether or not they will enjoy the video, but do not know the probabilities associated with those outcomes for each video on offer. As we learn more about climate change, we are moving from ignorance to ambiguity; the probabilities of different changes in climatic variables are becoming clearer, but the precise consequences in terms of species loss and water availability, for example, are not yet clear.
views, conclusions, and recommendations contained in the working paper are not to be taken to represent the views of the Commission
this approach is not possible when decisions must be made under uncertainty. we have no decision uncertainty about preferring B to A. the implication is that we should seek to maximise the coping or adaptive capacity of those individuals or communities. during and after all floods rather than simply construct an engineering solution that protects up to some design standard flood. conclusions. such as to past flood losses and past return periods. the appropriate response is to take an adaptive management approach: to treat all actions as learning experiments. The flood management strategy must include management of more extreme floods and also what to do when one or more elements of the adopted strategy fails.World Commission on Dams Assessment of Flood Control and Management Options 43 Methods for exploring the implications of decisions under risk are well-established. so allowing time to respond to the changing risk. Moreover. Such outlines suggest that we are both certain as to the extent of the future 100 year flood and that there is something qualitatively different about the flood risk within and outside of the 100 year flood contour. decision uncertainty does not exist as long as we are quite confident as to the rank order of preference across the two best options: rank order is all that is important and we need not consider further the remaining options once we are certain as which are the two best options. 4. Unfortunately. if there were only two options and we are very uncertain about the outcomes but are certain that whatever the magnitude of the outcomes. One reason why we may be uncertain what to do is because we are uncertain about the outcomes of the different available options but we can be undecided for other reasons as well (Green et al 2000). Holling (1978) and Walters (1986. It is no use concentrating our understanding upon refining our knowledge of the past. or decisions. then only in the most trivial sense is there still a choice to be made. ambiguity or ignorance. they will be worse under option A than option B. may be argued to inherently involve decision uncertainty in that if one is certain as which option to adopt.1 Managing all Floods and Not Just Some The flood management strategy adopted must cover all floods and not just some floods. Choices. 1997) have proposed that given the inherent uncertainty of decisions. Conversely. Gunderson 1999). the approach should be to look for resilient systems (Holling 1973. not just those up to some design standard of protection. we should seek for ‘forgiving’ systems: those that fail gracefully. which lack the same selforganising capacity as natural systems. Secondly. With artificial systems. that with natural systems. if the future is likely to be different. This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . it can be that we can be very uncertain about some of the outcomes of a decision without being very uncertain about what to do. it is necessary to plan how to respond before. we can propagate probabilities in component systems through Monte Carlo modelling and similar techniques in order to obtain probability functions for outcomes (Greeley-Polhemus Group 1992). That it. The ideal flood to manage is one: • which has a slow rise and in generally changes slowly over time. and • where there are no sudden changes in the nature of the appropriate coping strategy. The process of choosing is thus one of reducing decision uncertainty. drawing 100 year flood outlines on maps is both misleading and usually inaccurate. when dealing with individuals and communities. The views. Being precise when are we uncertain can be also be dangerous: for example. However. Finally. For instance. That decisions are about the future and must be made under conditions of uncertainty must be recognised and built into the management approach adopted. the important form of uncertainty is decision uncertainty so that being uncertain may be defined as a ‘state of rational doubt as to what to do’ (Green et al 2000).
conclusions. the more likely it is that that action will be successful. For example. dike systems may be designed with weak points so that failure is likely to occur at the most manageable point. including those floods for which the primary management strategy will fail. not least because they tend to expect the future to be like the past. Time is a critical resource in managing a flood. of bridge. But for more extreme floods this may no longer be a safe option and the appropriate option may be to evacuate to higher ground. The slower the rate of rise of a flood. It is necessary to design for failure: in developing the strategy. Sudden changes of state such as the breach of an embankment or the river taking a new course are difficult to cope with. Ideally. Thus. discontinuities in the challenge presented by a flood are particular problems (Table 14). dam): causes flood wave to travel downstream at high velocity Breach or overtopping of dike Velocity exceeds 2 metres/second: probability of building collapse is significantly higher above this threshold Depth exceeds 2 metres: probability of building collapse is significantly higher above this threshold It is therefore essential to plan how to manage all floods and not just some floods. road or railway): area behind floods when ridge line is overtopped or breaches Failure of artificial throttle points (e. the response which is successful for some floods would be successful for floods of all magnitudes. the greater is the time available to mobilise resources.World Commission on Dams Assessment of Flood Control and Management Options 44 Thus. people may have learnt to sit out a flood in their home or in another place of refuge. flash floods generally present a much more intractable management problem than mature rivers. In general. gorge): when flood flow reaches the discharge capacity of the gorge. creates upstream reservoir with a significant depth of flooding Depressions that flood to a significant depth when the rim is overtopped Natural ridge line or artificial embankment (e.g. all actions take time.g. Thus. plan future response and implement that response. bridge): when flood flow reaches discharge capacity or throttle point is created by debris. explicit consideration should be given to the mode and consequences of failure. including deciding what to do. This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. the longer is the time available to take action. Both individuals and institutions learn from past experience and apply that learning to future events but that learning is not always appropriate to all floods. a virtue of a channel improvement scheme is that it reduces the amount of water out of bank in all floods and not just those up to the design standard flood (MAFF 1999). The views. an upstream reservoir is created with a significant depth of flooding Artificial throttle points (e.g.g. communicate knowledge. Flood management should not be limited to the assessment of the effectiveness of a proposed strategy up to some essentially arbitrary ‘design standard’ flood but the impact of that option on the difficulties of managing all floods must be considered. Thus. and how those will be managed. This is seldom the case and those responses which are successful for the more frequent floods are often very dangerous ones to adopt for more extreme floods. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . Table 14 Potential forms and causes of discontinuities in the challenge presented by a flood (Green 2000) Permanent change in course of the river Natural throttle points (e.
through This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. a flood in a catchment arises from an event and that event may result in floods across several catchments. • as a characteristic of the flood to which that population is exposed. Whilst the probability of a more extreme flood occurring in that area is indicated by the return period of the design standard event. Blaikie et. al. or in a poorly protected coastal zone in an otherwise affluent European setting. and ability to mobilise. number and strength of conflicting groups). it includes adaptation to floods and response during a flood.g.partly reflecting different disciplinary and ideological positions. defining vulnerability as: ‘the characteristics of a person or group in terms of their capacity to anticipate. morale. or • as the interaction between the nature of the flood and the characteristics of the population at risk. economic and political processes and differential access to. the subject of much debate and there are a variety of definitions . ‘Resilience’. The causes of vulnerability lie in social. rivalries. The components of vulnerability have been analysed by Cannon (2000) and include (a) initial well-being (e. or designing to.g. severe floods in Bangladesh are the result of the three great rivers being in flood simultaneously. assets and savings (c) selfprotection (e. however. If there are fifty catchments in a country and the probability of a flood in one is independent of the probability of a flood occurring in other. capacity for self-reliance) (b) livelihood resilience (e. as for (a) plus building regulations. It can therefore be more useful to think in terms of the probability of an event rather than the probability that a particular river will be in flood. and recover from the impact of a natural hazard’. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . it says nothing about the probability of an extreme event occurring somewhere in the country. then each year the chance of the 500 year return period flood occurring somewhere is 1 in 11. The concept of vulnerability is. livelihood type. Conversely.g.or under-insured and living in ignorance of the flood risk in one-storey retirement home or mobile homes within a floodplain. vulnerability to hazards such as floods is often made worse by poverty and through related weak response and recovery capacity. but not always. They can also include elderly. low-income pensioners who are un. (1994) took the first approach. physical and mental health. which may be taken as the opposite of vulnerability. thus. cope with. resources.g. High vulnerability frequently promotes and exacerbates flood disasters and is often. rural Malays to floods in eastern Peninsular Malaysia is reduced by close kinship systems which exist in the floodplain villages.g. infirm and the immobile. Vulnerable people can be rural-urban migrants living in poverty in squatter settlements in Manila in the Philippines. Although vulnerability and poverty are not synonymous.2 What is Vulnerability and Who is Vulnerable ? The concept of ‘vulnerability’ is a central one in understanding the effects of floods and the means required to alleviate these effects. social cohesion. Vulnerability is an on-going state or condition in which people exist rather than a status that can be associated with any particular hazardous event. a single design standard flood will engender a false sense of security. may be enhanced by promoting access to knowledge and resources. building quality. and different purposes. Whilst this definition of vulnerability includes recovery after a flood and capacity to recover. income opportunities. conclusions.World Commission on Dams Assessment of Flood Control and Management Options 45 Planning for. nutritional status. Chan (1995) demonstrates how the vulnerability of poor. There are three main definitions of vulnerability: • as a characteristic of the population at risk. Such a risk should not just simply be ignored. The views. or the landless labourers and families living in flimsy shelters in low-lying areas of Bangladesh. Vulnerability may then be reduced in a variety of ways. Vulnerable people may include the elderly. technical interventions) and (e) social capital (e. resist. qualifications. protection from hazards. 4. location of home and livelihood) (d) societal protection (e. closely associated with under-development.
In such studies. Vulnerability is viewed as a product of the interaction of the flood event and some of the characteristics of the affected land uses and population. This third form of definition leads to a form of ‘vulnerability analysis’ which has become an all-encompassing means of assessing risk and social and economic consequences associated with a hazard (Alexander 1993). reducing dependence and increasing the success of the population in mobilising resources. in better healthcare. the risk of buildings being structurally unsound. conclusions. as that which results when coping resources are inadequate to meet the challenge posed by the flood. Using this interpretation vulnerability can be determined by a hydrological analysis only. The views. speed of onset etc. ill-advised interventions to attack a perceived problem may increase the vulnerability of some groups. In the longer run. The weakness of the first definition is that implies that a community will have a similar level of vulnerability whatever the nature of the flood event. and the effects of an event in terms of estimated property loss and casualties are all measurable and combined to derive a technical estimate of ‘vulnerability’. 4.2.World Commission on Dams Assessment of Flood Control and Management Options 46 poverty-reduction programmes and through processes of development as reflected. for example. estimates of flood probability. and the characteristics of the population and land use systems affected are either ignored or are only considered as peripheral matters. the floods magnitude. in particular. and recommendations contained in the working paper are not to be taken to represent the views of the Commission .g. The third definition of vulnerability is akin to Lazarus’s (1966) definition of stress. Conversely. Penning-Rowsell (Green et al 1994) proposed that the following equation be used to define vulnerability of households to flooding: This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. Adams (1992) has argued that this has been the consequence of a number of dam-based projects in Africa. Development projects must however aim at improving self-reliance. housing and infrastructure. Thomas and Adams (1999) have shown in one case that adaptation to the changed conditions was more successful than in the short-run but the success of this adaptation may have been due to exogenous factors which counterbalanced in part the problems resulting from the original project.1 Alternative definitions of vulnerability The second definition of vulnerability to flooding is simply and solely interpreted as a function of the characteristics of flooding (e.). This definition clearly only implies that there is a flood problem which should be ‘fixed’.
R . W t. the choice of projects might well involve strategies designed to reduce the population’s vulnerability rather than providing some kind of direct flood alleviation. 4. or to train the members of vulnerable communities in how to prepare for floods and to spread a preventive culture (Parker 2000). W a . Sd. coastal or not) Velocity Pollution load of flood waters Rate of water rise during flooding onset Whether a flood warning received Warning time provided Advice content of warning was Susceptibility of contents to damage Susceptibility of building fabric Time taken to restore Wo infrastructure Number of storeys Wt Robustness of building fabric Wa T Ra Rq = = = = = = Time taken for assistance to arrive after or during event Amount of response available Response quality Vulnerability in the forgoing discussion is implicitly defined at the level of the household or community. Ra. the result is that vulnerability assessments yield a series of overlays. I.St. Such projects might involve decentralising decision-making and integrating the flood knowledge and coping mechanisms of local people with those of flood agencies. Ro .World Commission on Dams Assessment of Flood Control and Management Options Property and infrastructure variables 47 Household vulnerabiliilty = Social/economic variables Flood characteristics Warning variables Response variables = f A H. appropriate projects might seek to diversify and strengthen fragile local economies which occupy flood-prone areas. However. Green 2000). The views. different elements or systems being vulnerable to different extents to different characteristics of floods. Indeed. This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. F .1. Pj. W o . V.2. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . Rq where: A H S I C F Sc Sb It St Ro = = = = = = = = = = = Age profile of household De Health status and/or mobility of Dt household Savings of household Sd Household income St Cohesiveness of community W Flood knowledge V Pl building R = = = = = = = = Depth of flooding Duration of flooding Sediment concentration Sediment size Wave/wind action (e. Sb. Sc. Tr. De. Dt. conclusions. S.1 Implications of Interpretations of Vulnerability for the Choice of Hazard Alleviation Projects Flood Under the first interpretation of the concept of vulnerability explained above. Such vulnerability assessments have the potential to be used to screen large areas of a country in terms of the priority that should be given to undertaking more detailed flood management assessments (e.g. C. ecosystems and national or regional economies may also be assessed in terms of their vulnerability. flooding here is interpreted essentially only as a symptom and not as a diagnosis. It St.g. Such assessments need to be made using different parameters. Alternatively. In some cases flood alleviation projects might be combined with such measures. W.
such maps will still contain little useful information about the threat posed by a flood unless they indicate likely depths of flooding and areas where flood velocities will be high. the 500 year flood is going to be significantly different from the 100 year flood. such as road and railway embankments. For management purposes it is the major changes in flood routes in particular that should be identified: for example. In this case. but the analysis should be This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. Those features of the floodplain which will limit flood flows and extent should also be identified. then it will simply be necessary to show where the most vulnerability segments of the population are located. such models and maps therefore implicitly assume the second definition of vulnerability because they exclude all of the characteristics of the populations exposed to the flood. This will show both critical flood features and population characteristics. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . Again. For this reason. the flood ways should be identified. Constrictions on flows such as bridges should also be shown (Figure 8). Although the information will necessarily incomplete. Conventionally. The objective is to manage all floods and not just some. particularly out of bank flood routes. since they contain no information about the threat posed by more extreme floods. The third definition of vulnerability is likely to be more useful in terms of management. The third interpretation of vulnerability which stresses interaction between the flood and the population leads logically towards projects which seek to reduce the challenge presented by the flood and/or to enhance the resources available to people to cope with the event when it occurs. and areas of passive and active storage.World Commission on Dams Assessment of Flood Control and Management Options 48 Under the second narrow interpretation of flood vulnerability (see 4. On the flood side. a different form of modelling and mapping is necessary. The simplest form of such a model is a floodplain map. as well as natural depressions should be indicated. and perhaps of different durations or seasonal characteristics. both depths and velocities are likely to be crucial. retard flood generation and so on.1 above). indeed. conclusions. or may range up to maps showing in accurate detail the extent. But the problem with this approach is that it may not address the fundamental socio-economic and political causes of flood vulnerability. A second typical limitation of the simpler maps is that they simply draw a boundary around the flood extent for some flood of historic record or the modelled extent for a flood of a particular return period. drawing a boundary to the flood extent is something that should be avoided. 4. This is not so if the flood management specialist is presented with a set of floods of different return periods. they are static. Thus. perhaps an old meander belt. buildings and population result in a high degree of vulnerability. if the 500 year return period event will cut across a low area. Thus. banding or zoning is more appropriate than lines. produced from comprehensive mathematical or physical models. This requires that the development of the flood waves be modelled. These approaches give picture of differing levels of accuracy and detail of the flooding hazard at a single moment in time of a historical or idealised “design” flood.2. they ignore the inherent uncertainties. based on comprehensive mathematical models To be useful for flood management. such as the 100 year return period flood. which assumes that flooding mechanisms are the principal problem to be addressed. velocity and depth of flooding for defined return periods. an important datum for flood management is whether. because of all the uncertainties. This may be based on an approximate interpretation of the records of the largest floods of historic record. focusing on a single event. projects are likely to be selected which seek to reduce flood magnitude. both natural and artificial crestlines. for example. Even so. should be avoided. now occupied by urban development. If the first definition of vulnerability is adopted.3 Identifying the threat The necessary first step to flood management is to identify the nature and extent of the threat. and do not observe the principle that it is necessary to manage all floods and not just some. The views. and because of the necessary uncertainties. Although the length of record and other factors will limit the reliability with which the forms of floods of different return periods can be modelled. it is necessary to indicate those areas where the combination of flood characteristics.
2m passive storage: depth < 1 metre floodway bypass if flood level exceeds 15. points where emergency flood fighting works may protect the areas behind them.World Commission on Dams Assessment of Flood Control and Management Options 49 taken further to indicate areas where structures are likely to fail to provide protection to their occupiers. where buildings should be discouraged if possible because they lie in a flood way. may also be indicated. Critical installations and those giving rise to potential secondary hazard. such as areas where toxic or flammable chemicals are stored. and where flood proofing of buildings may be satisfactory because they are only in a passive flood storage area.3m velocity >4 m/sec depth <1 metre camp site: depth c 2 metres The purpose of such a map in its simplest form. is to inform flood management. This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. The views. 7 Figure 8 Hypothetical flood hazard map active flood storage depth < 1 metre factory storing 200 tonnes LPG railway bridge geriatric hospital railway embankment: level 16. those areas where warning is necessary. conclusions. which areas should be evacuated in the event of a flood. A “layer” on the map then may be used to identify the appropriate management responses: those points such as bridges and crest lines which need to be watched. and recommendations contained in the working paper are not to be taken to represent the views of the Commission .
as shown by their impact on the flood hydrograph (Figure 10). or probabilities. Again. A problem in operating such storage is that effective operation requires reliable flood forecasts. flood warnings are necessary if such strategies as contingent flood proofing and evacuation are to be successfully achieved. flood warning sits at the centre of the matrix of options.World Commission on Dams Assessment of Flood Control and Management Options 50 This is only one potential purpose for a flood hazard map. households and communities must confront or to enhance their coping capacity. In France. This reduces the runoff that reaches the river but may be less effective at the height of the rainfall event. the dam may have been constructed with insufficient spillway capacity to safely pass an extreme flood and sluices may have to be opened in order to protect the dam from failure. particularly if one storm event is closely followed by another. 4. There are a large number of newspaper and other reports that assert that flooding was made worse by releases from a dam. or standard river engineering options. if a flood peak is followed by an unpredicted flood peak then insufficient water may have been released to store the new peak. so increasing the forecast lead time. Similarly. At the same time. Finally. or that is the theory. Strategic storage. maps define areas where no building is to be permitted and those where only flood proofed buildings are to be permitted. the flood downstream of a dam should under all circumstances be no worse than if there were no dam. is usually operated to reduce the peak flows and water levels. the water so stored being released later on during the falling flood or recession. as well as changes in the relationship between flows and levels in vulnerable areas. but these reports give insufficient detail This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. The chance of managing a flood through mobilising resources is increased if it develops slowly. the flood crest is both slowed and lowered but lasts longer. having both physical elements in the form of raingauges. In theory. this will be true provided that the original hydrological analysis and subsequent flood forecasts are reliable. This differentiation is somewhat different from the usual split between ‘structural’. Figure 11 is represents a simple case. In practice. effectively one where a single catchment or sub-catchment generates the flood wave. there is no storage available to absorb that peak. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . conclusions. may be possible in urban areas and has been much promoted. of occurrence. Hydrological analysis will identify the flood flows or flood levels associated with particular return periods. these interventions may be physical in form or involve institutional changes. if the hydrological modelling of the catchment was inadequate when the dam was built. for example. weather radar and streamflow gauging stations and an institutional element which converts the flood forecasts derive through that system into useful flood warnings. As shown in Figure 9. flood management strategies can be categorised as either seeking to reduce the challenge the individuals. At the other end of the spectrum the steep flashy river in mountainous or arid areas presents the greatest challenge in this respect.4 Reducing the Challenge Under the Lazarus model. Such reservoirs operated for flood control are usually kept at a low level to absorb incoming floodwater. Therefore. essentially involving many small units of local storage at source. involving dams large or small. In more critical cases possible variations in precipitation and runoff over time as a result of changes in climatic change and land use will increasingly have to be taken into account. because the local storage may become saturated. The views. The different management options available (not all will be feasible in the local circumstances) change the challenge presented by the flood flow in different ways. If flood insurance is available then a ratings map defining areas at different levels of risk is also required. Indeed. and ‘non-structural’ options which essentially cover everything else. the likely reliability and lead time that can realistically be expected of a flood forecasting and warning system can significantly influence the choice of options to be adopted. if a forecast is unreliable then the storage may be used to absorb to what is forecast to be the peak. The latter in particular requires a range of flood events of different return periods to be modelled. Source control. If the flood then goes on rising.
conclusions. The effect of channel improvements. the absence of floods has been known to be attributed to the presence of dams that have no intended flood alleviation role and equally it is to be expected that an unusually extreme flood will be blamed on any upstream dam. Figure 9 Changes to flood hydrograph by management option This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. In addition. the operators were not prepared to draw down the reservoir in advance of the flood so as to allow sufficient storage for the flood wave. A final potential explanation is that the release of water resulted in a much faster rate of rise in flood levels than experience with ‘natural’ floods had lead downstream residents to expect. The views. the depth of flooding will always be less with house raising than the depth of flooding now . although in areas like the River Tone in Japan where maximum flows are hundreds of times greater than minimum flows. the losses from all floods will be less than under the current conditions. as rational animals we all seek explanations for both the absence of floods and the occurrence of floods.World Commission on Dams Assessment of Flood Control and Management Options 51 to assess where the problem lies. this difference may be negligible. Figure 11 shows how the extent of the economic or financial losses are likely to vary according to the extremity of the flood. One explanation is that flood forecasts were erroneous resulting in operational failure. But the detailed analyses required to distinguish between these potential explanations do not appear to be available. For house raising. losses will always be less than they are now. The fourth explanation of the reported problems lies in human psychology: those living downstream have adapted their behaviour to the reduced flood experience and are consequently more exposed to those floods that do occur. the building of embankments or flood diversion channels is simply to raise the flood flow which can be passed without flooding resulting. For both channel improvements and house raising projects.unless flow velocities in some floods are such as to demolish the building. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . a second is that with multi-purpose dams. In the case of channel improvements this is because there will always be less water out of bank with the project than there is now. The third explanation is then that the spillway design was based an underestimate of the magnitude of the extreme flood and emergency releases of water took place.
Difficulties in evacuating the flood water after the breach has been closed may also add to flood losses. as soon as this happens. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . The views. With dikes. Thus there will be zero damage up to the design flow. it makes little difference whether is the 100 or 200 year return period flood that causes the flood. the line is shown as vertical rather than sloping: once a dike has failed. and much reduced damage above this as excess floodwater is spilled into the river downstream. losses are reduced to zero until the dike overtops. losses are reduced to zero until flood levels exceed the standard of protection offered by flood proofing. Dams are usually governed by strict national legislation covering their design and inspection. with failure rates many orders of magnitude less than river embankments. Dams protect against all floods up to a certain volume and even larger floods will be attenuated and slowed. losses will be identical to those which would be experienced in the same flood if it occurred now. Such failures are extremely rare and fall outside the scope of flood protection – so that strictly the curve for dams should be well to the left of the curve for dikes and in theory lie on the y axis. Figure 10 Schematic differences in the loss-probability curve by management option This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. Since overtopping can often lead to dike failure.World Commission on Dams Assessment of Flood Control and Management Options 52 hydraulic impacts of management options 12 10 8 flow 6 4 2 0 11 13 15 17 19 time 21 1 3 5 7 9 base flow source control storage slow protect With flood proofing. conclusions.
1) and the risk of flooding will increase markedly. The change of risk as a flood develops is given by the appropriate segment of the line for that management option from its right hand extreme to the vertical line that indicates the return period of the peak of the flood. the diagram also shows how the risk to life changes as a particular flood develops. However. With house raising or flood proofing. the risk to life will be equal This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. it being the shape of the curves that is important. this will no longer be the case and some buildings will collapse or be swept away. At this point. conclusions. At present. the depth of flooding and/or its velocity will exceed a critical threshold (Section 3. For clarity. and thus that of a particular return period. Whilst the diagram is shown as a plot of the risk to life versus the exceedance probability of the flood event. the dwelling provides a safe place of refuge up to some severity of flooding. the without project condition. the risk to life increases slowly as the probability of the flood decreases (and consequently its magnitude increases). particularly when the velocity and/or depth is sufficient to cause the collapse of buildings. In each case. The views. the scales are distorted. At some flood magnitude. when flood levels or velocities reach specific levels. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . the risk to life is formally defined as the risk to life conditional on the flood management option adopted and excludes the possible reduction in risk that might be achieved by a flood warning and evacuation. The locations of each curve are also determined by the particular design standard of protection offered by the approach in the individual local circumstances.World Commission on Dams Assessment of Flood Control and Management Options 53 Loss losses with dikes or dams losses with channel improvement project if drowned out losses with channel improvement or house raising project losses without project losses with flood proofing design standard of protection return period Figure 12 then compares the likely risk to life under the current conditions and those prevailing if different management options are adopted.
exceed 30 metres in height. the dike should only fail when it is overtopped but in practice. there is a risk to life from dams under other conditions as a result of so-called ‘dry weather’ failure. that in extreme floods. This probability is likely to be considerably higher than from any natural occurring floods. However. Because these curves are defined in terms of the risk to an individual. the risk to life is zero until the dike fails. it assumed that above some rate of inflow to the dam an increasing proportion of that inflow is discharged. The views. the exceedance probability of the PMF is very close to zero. the risk to life rises nearly instantaneously. then there will always be less water out of bank than under current conditions whatever is the return period of the flood. with a suggested probability of 0. In addition. Where there is a significant number of people living below the dam then the spillway will generally be designed to be able to discharge the Probable Maximum Flood (PMF). The risk to life from a catastrophic dam failure attentuates with distance downstream from the dam as the energy of the flood wave is lost and usually the risk to life has fallen markedly at a distance 25 kilometres below the dam (and a lesser distance in the case of smaller dams). and recommendations contained in the working paper are not to be taken to represent the views of the Commission . Furthermore.00. At this point. When the dike fails either as a result of overtopping or breaching. In principle. in some cases. However. if the capacity of the river channel is increased. if the dam fails as a result of an extreme flood. they do not change as a result of any development that is attracted to the area after the adoption of one or another flood management option. the flood storage capacity of the dam is exhausted. With a dike. the severity of flooding with the dam is identical to that prevailing under the without project condition. dam spillways are designed to have capacity to discharge inflows of greater or lesser return periods. within the area immediately downstream of the dam. at this point. and unlike the other flood management options. the risk curve is below and to the left of the curve under the without project conditions. Consequently. Graham’s analysis of past floods leads him to conclude that the probability of death will be in the range of 0. For the dam option. Graham (1999) notes that the flood wave after catastrophic failure could. the discharge flow from the dam equals the inflow and the dam ceases to have any affect in attentuating flooding. by that time or earlier.World Commission on Dams Assessment of Flood Control and Management Options 54 to the risk under the without project conditions. there is a risk that the dike will fail by breaching because it has been poorly constructed or maintained (Wolff 1997). The velocity of flow through a breach in particular is likely to be greater than the velocity that would exist under the without project condition and consequently the risk to life under the with dike condition is shown as being higher than without the project.75. theoretically. This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. conclusions. although that risk may be elevated during flood conditions that are not in themselves sufficient to cause failure through overtopping. Finally. In addition. with the possible exception of some flash floods in small catchments. Thus.30 to 1. the resulting flood immediately downstream will be catastrophic. Depending upon the extent of development downstream. the conditions outside the dwelling will be such as to make it impossible to escape unaided.
with the flood being somewhat attenuated on its passage down river. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . whilst the peak flow on the tributary is higher than on the mainstem at B. increasing storage on the main stem is quite problematic: to allow for the passage of the flood crest from the tributary. in both cases. an approach has been adopted which results in the same maximum flood flow at B. the total discharge from the mainstem is much greater than from the tributary. as well on the flood on the main stem. In this hypothetical instance. This pattern would result if. In the case where the first flood crest that arrives is that from the tributary. the flood flow at B depends in part on the flood flow from the river on which A is located. the flood crest from the tributary might reach B after that of the main stem (Figure 15). Alternatively. In the worst case. We might delay the crest from the main stem by controlled storage on the floodplain in the form of a detention basin. a large proportion of the flow on the main stem must be diverted to storage. The views. we might seek either to reduce or to delay either crest. with a short time to concentration and a short flood duration. Figure 16 illustrates the first approach and Figure 17 the latter approach. the two flood crests would coincide in their arrival time at B and this is a situation which it is obviously highly dangerous. Pumping will This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. In this instance. the flood on the main stem is characteristic of a mature river system with a slow rate of increase and long. Alternatively. flat flood crest. conclusions. Diverting the flow in this way uses the fact that the water level in the river channel is usually higher than that in the detention basin. we might construct a flood reservoir on the tributary to store the flood crest and discharge it after the flood crest from the main stem has passed B. for example.World Commission on Dams Assessment of Flood Control and Management Options Figure 11 risk to life risk with dams 55 Risk to life by management strategy and exceedance probability risk with dikes risk with house raising/ flood proofing risk without project risk with channel improvement project return period A more realistic situation is shown in Figure 13 where a river is fed by a number of tributaries. A flood from the tributary may arrive at B before the flood crest from the main tributary (Figure 14). Conversely. The flood crest at A shows the characteristics of a flood on a rather flashy catchment. floods are typically the result of rain front moving east to west.
conclusions. This example is intended only as an illustration of the problems of managing flood flows in complex river systems and it is not claimed that there is any universal applicability of the conclusions. a flood storage dam on the tributary (Figure 16). The views. In the alternative approach. Figure 12 Flood management in a hypothetical catchment A B This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. the flood crest on the tributary is stored and released only after the crest on the main stem has passed.World Commission on Dams Assessment of Flood Control and Management Options 56 usually be both unnecessary and uneconomic. and recommendations contained in the working paper are not to be taken to represent the views of the Commission .
The views. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . conclusions.World Commission on Dams Assessment of Flood Control and Management Options Figure 13 Tributary floods first: flood hydrograph at points A and B flood hydrograph 30 25 20 flow 15 10 5 0 11 13 15 17 19 21 time 23 1 3 5 7 9 tributary A tributary B mainstem B total B 57 Figure 14 Main stem floods first: flood hydrograph at points A and B flood hydrograph 25 20 flow 15 10 5 0 13 17 tim e 21 1 5 9 tributary A tributary B mainstem B total B This is a working paper prepared for the World Commission on Dams as part of its information gathering activities.
These processes reduce or eliminate sudden surges in the volume of surface water run off and the quantity of pollutants carried into rivers. conclusions. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . The views. thus restoring base flows. deforestation and other changes in land use higher up a catchment may reduce the ability of rain water to drain into the soil and increase the speed and volume of surface water run off into existing water courses. Source control aims to mitigate such problems by recovering the ‘natural hydrological cycle’ to the greatest extent practicable so that water is retained close to its source and allowed to infiltrate into the soil restoring ground water and through the soil into watercourses gradually.World Commission on Dams Assessment of Flood Control and Management Options Figure 15 58 Flood hydrograph at points A and B with flood storage on the main stem flood hydrograph 25 20 flow 15 10 5 0 time tributary A tributary B mainstem B total B Figure 16 Flood hydrograph at points A and B with flood storage on the tributary flood hydrograph 25 20 tributary A 15 flow 10 5 0 10 13 16 19 tim e 22 1 4 7 tributary B mainstem B total B 4. agricultural land drainage. This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. This can lead to flood problems lower down the catchment as well as increasing the flow of pollutants into rivers.4.1 Controlling runoff Urbanisation.
car parks and industrial sites may be covered with pollutants which can be washed off. 1999b). adopted a strategy involving carriageway reservoirs and percolation wells and drainage pits (Balades..4. Bordeaux with a considerable problem of flooding due to urban surface water which could not drain naturally into its river. conclusions. • small extended detention basins: grassy and vegetated depressions to hold and treat excess surface water for slow release. a series of publications on sustainable urban drainage systems or urban source control are in preparation for government departments and other agencies by the Construction Industries Research and Information Association (CIRIA). US studies have shown that downspout diversion programmes can reduce mean flow volumes in the sanitary sewer network by over 25% with flow reductions ranging from 25% to 62% for rainfall depths of between 6mm and 25mm (Kaufman and Wurtz. usually located between paved areas and the watercourse to slow flows and remove pollutants. • retention ponds or permanently wet ponds which retain surface run off and provide biological treatment through wetland and aquatic vegetation such as reeds. This is a working paper prepared for the World Commission on Dams as part of its information gathering activities.World Commission on Dams Assessment of Flood Control and Management Options 4. amongst other countries. a companion design manual for England and Wales and a review of UK. 1992). The types of techniques recommended include: • infiltration trenches and soakaways. demonstration sites and best practice in source control (CIRIA1999c). The use of infiltration techniques is only feasible where this would not pose a threat to ground water supplies and where there is no question of land contamination. These include a design manual for Scotland and Northern Ireland (CIRIA. In the UK. and carry it off site to the nearest watercourse as quickly as possible. infiltration and treatment systems which can also serve as landscape features. as used in some UK and French motorway service areas and supermarkets : • swales or grassed surface conveyance. Impermeable surfaces such as roads. et al. In the USA. Source control in urban developments aims to control the quantity and quality of surface water runoff and may also enhance the amenity and wildlife value of a site. allowing domestic downspouts to discharge to lawns or paved areas instead of being connected to the sewers) has helped to maintain a consistent flow of higher water quality into urban streams. • wide filter or buffer strips of any natural vegetated form: grass or woodland. • permeable or porous pavements for roads and car parks. direct run off can increase to over 80% of the volume of rainfall (CIRIA. only 15-20% of the volume of rainfall becomes direct surface runoff that drains slowly into watercourses. car parks etc. recharge of regional groundwater supplies from downspout diversion programmes (i. 1999a).e.1. 1997). The views.1 Urbanisation 59 On undeveloped sites. and the use of conventional piped drainage systems which were aimed to collect surface water from roofs. With development with buildings and paved impermeable surfaces. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . • infiltration basins which hold surface water allowing it to infiltrate the soil gradually.
However. Finally. natural wetlands are important assets for flood storage (Maltby 1986) and artificial wetlands are being created for this purpose.World Commission on Dams Assessment of Flood Control and Management Options 4.4. This process has been the subject of intensive study with mixed results. pesticides and agricultural waste products.1.2 Agricultural land drainage 60 Where cultivated land adjacent to rivers has been drained to permit more intensive cultivation of higher valued crops to take place. 4. semi-permeable or wholly impermeable. Dams can be of several kinds. or in the lowlands as a detention basin which is normally largely dry except when required for flood storage. The UK Forest Enterprise and Washington State in the US have policies designed to counter the illeffects of forestry. In some cases. an important example being the Dongting Lake in Hunan Province. Outside of the flood period. Nepal has a national programme of reforestation. dams may fail if the flood is extreme and some forms of dam can fail very rapidly.4. the lakes associated with a river can be used for detention.2 Storage Source control can itself be considered a form of storage but the storage of the flow in rivers can take place either in the uplands.3 Deforestation.1. retention ponds and wetland areas are other ways in which agricultural run off can be retained and treated. forthcoming). There is evidence.4. channel siltation and landslides caused by inappropriate clearfelling and logging techniques exacerbate the flood problems. where the high dam and associated reservoir is the traditional form. waters flow normally through the wire mesh if the debris has been removed. the difficulty of providing transferable predictive models on the basis of these results indicates that individual catchment characteristics matter and that many experiments have been poorly designed (Jones. Unplanned dams are a particular threat in this regard: bridges may be blocked by debris and then fail as the water level rises on the upstream This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. Such floods sweep the ground clear in their path until attenuated. vegetated buffer strips between cultivated land and watercourses can be used to slow surface water runoff and treat the water to remove pollutants (Delaney 1995). in California. Bosch and Hewlett (1992) on the basis of world-wide experimental results concluded that deforestation tends to increase run off and peak floods at least in the short term. conclusions. Inflatable dams have also been used. reforestation and afforestation Deforestation and the removal of natural vegetation in the upper parts of rivers has been held responsible for flooding problems lower down the catchments in many parts of the world. however. 4. careful management practices and riparian buffer zones are possible. However. the flood borne debris is trapped in the wire mesh and blocks much of the flood flow. boulders have been used to create artificial rock pools which capture and store the rising crest of a flood. Possible methods for mitigating these risks are to adopt less intensive agricultural practices and to control water levels to ensure that water is retained on the land. These may also provide features for wildlife. that afforestation normally results in increased evapotranspiration but does not necessarily reduce the flood risk. The views. Similarly. For example. in Trinidad wire mesh dams have been used in the highlands: during flash floods. semi-permeable dams were used to capture flash floods to provide irrigation water for the subsequent growing season. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . principally. infiltration ponds. this has usually increased the volume and speed with which rain falling on cultivated land has reached watercourses with the potential to raise the risk of flooding downstream and the possibility of pollution of watercourses by herbicides. China. Chan and Parker (1996) suggest that increased soil erosion. In pre-Roman North Africa. The consequence can be a catastrophic flood with a wall of water metres high moving at high speed. Programmes to mitigate possible impacts of deforestation through reforestation.
In urban areas. the people living downstream will have become accustomed to the new flood regime produced by the dam and will be unaware of the remaining risk. The second is that even with sophisticated flood forecasting systems it is impossible to know the future with any certainty. farmers being paid for the damage to their crops when the basins are used. as well as storing flood waters. the water stored must be released at some point. conclusions. There are a number of problems in operating dams for flood storage purposes. Thus.000 . The first is that dams are rarely used for flood storage alone and the optimum strategy for flood storage and these other purposes can conflict. But. not being utilised too early and discharges also being controlled. in tropical zones. Secondly. there will be no capacity left to store part of a second and perhaps larger crest.000 per km2 in capital costs plus an additional $US 5. where it is required that one of the options considered when the renovation of an existing flood alleviation scheme is being appraised is the managed retreat option. we are increasingly re-introducing them artificially: those along the Rhine being an example. Wetlands. some 1. the reservoir should be in a drawn down state. frequently also slow it down. it then has to be evacuated after the flood crest has passed.World Commission on Dams Assessment of Flood Control and Management Options 61 side of the ‘dam’. detention areas are often used for sports fields. Detention basins are usually quite shallow and hence a large area is required in order to provide flood storage. Thus. the large number of detention basins are all occupied so that the detention basins can only be used after the population is first evacuated. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . Thirdly. However. for hydroelectric generation. the option proved to be uneconomic. Natural wetlands can be excellent forms of detention basins (Delaney 1995) and artificial wetlands are also increasingly being constructed to store flood waters. Thus. they can be put to use in the meantime. the peak of which will in any case not be predictable with certainty. Where wetlands were engineered away in earlier times. it is generally held that storage has the greatest effect the nearer upstream it is to the area that it is sought to protect.000 km2 in annual compensation costs.$US 20. the ideal is to keep the reservoir full but to provide the maximum protection against flooding. Another option considered for the Red River catchment in Canada and the United States was that of ‘micro-storage’. Finally. attenuating the flood wave. in the UK. Quite commonly the method of allowing flood waters into a detention basins is simply to blow a breach in the protective dike: some dikes are constructed with demolition chambers in place. because they are normally dry and in most years will not be required. In general. Methods offering greater control are sluices which can be opened to allow varying rates of flow into the detention basin or sections of dike designed to overtop without failure. In that instance. evaporative losses may also be significant (Acreman and Hollis 1996). For instance.800 km2 of land being necessary at an estimated cost of nearly $US 100. Once water has been allowed into a detention basin. The views. using the agricultural fields between the raised roads as flood storage (International Red River Basin Task Force 2000). Along the Yangtze. A number of factors are important in the choice between the different possible forms of storage. A critical limiting factor is the availability of a site for the particular form of storage. Wetlands are also highly valuable resources for biodiversity as well as fulfilling other functions (Maltby 1986). storage is most effective if it is controlled. Dams may also fail for reasons other than extreme flood flows (‘dry weather’ failure). the detention basins of the Lincoln flood alleviation scheme are farmed (Parker et al 1987). the rate of discharge from the reservoir will not be greater than if there was no reservoir in the first place. The limitation of natural wetlands is then that they typically flood This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. if the reservoir is filled completely on the rising crest. This approach has the limitation that it is difficult to control the rate at which the detention basin fills. some farmland is likely to be of greater value if it is allowed to revert to its original wetland condition than in agricultural use (MAFF 1999).
conclusions. the storage capacity of the live stream is also reduced.4. on parts of the Rhine. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . the meanders are being restored. The argument against making artificial shortcuts or allowing the river to take a shortcut in a flood is three-fold. the time taken for the flood peak to travel from Basle to Karlsruhe on the Rhine having halved between 1955 and 1977: in consequence. In consequence. if the flood continues to rise then there is no remaining storage capacity. The views. the risk that the flood peak from the Rhine will coincide with that from the Neckar. the loss of instream storage speeds the passage of the flood peak downstream. there are economies of scale. the trend is in the other direction. the gradient will be steepened and so flow velocity and flood flows are increased. Firstly. Thus. again the ability to control inflows and outflows to storage are important to making the best use of that storage capacity. Control of the discharge so as to lop the peak of the flood flow are critical points downstream is also generally more difficult with a number of small storage areas (International Red River Basin Task Force 2000). Areas within a meander may be protected by dikes in order to prevent the river cutting a short route during a flood. areas of shallow and low moving flows are more valuable than areas of deep and fast moving flows. Nahe and Moselle has increased (Landerarbeitsgemeinschaft Wasser 1995). if the river length is reduced. In part this was undertaken in order to improve the river for navigation. although in most cases where the river is being restored to a more natural form (Brookes and Shields 1996) this is for environmental rather than primarily for hydraulic reasons (Brookes 1990).3 hectares) in size.World Commission on Dams Assessment of Flood Control and Management Options 62 early as the flood rises. the Danube and the Mississippi being well known examples. restoring meanders in some cases and avoiding shortening the length of a river. the Landerarbeitsgemeinschaft Wasser (1995) estimated cost DM10 per cubic metre of storage for large retention basins. The second argument is environmental. Fourthly. natural wetlands are likely to have most effect in ameliorating high frequency floods and to have much less effect on extreme floods (SAST 1994. Now. in general. In addition. For example. 4. Shulz (1999) reports that the cost of wetland restoration or creation falls from $US 300 per acre for small wetlands to $US 100 per acre for wetlands over 5 acres (12. rising to DM50 per cubic metre of storage for large retention basins and experience in France is similar. there is a 50 kilometre meander on the Yangtze in which the live storage is approximately 2 billion cubic metres. Shultz 1999).3 Slowing the flood wave In the past it was quite common to shorten the river by cutting off meanders and shortening the length of the river. This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. Thirdly. A further limitation is that the worst floods are frequently those that occur after an earlier event has saturated the ground and left local storage full.
conclusions.4 Carrying the flow A widely adopted approach has been to increase the capacity of the river channel by widening. This policy was reinforced by successive floods and the perceived need to control them with levees. deepening. Some rivers are being restored to a more natural form (ARSIA/GEOPLAN/CEHIDRO/WASSER WIRTSCHAFT 1999). An early example is the diversion channel built in 1621for the river Tone to reduce the flood risk to Tokyo. and by 155 levees extended along 1. More recently. In other cases. straightening.World Commission on Dams Assessment of Flood Control and Management Options 63 4.4. one example being the River Lea to the east of London. The Mississippi ‘Delta Survey report’ published in 1861 recommended a ‘levee-only’ approach to floods – a strategy pursued well into the twentieth century leading to more and more levees (Barry 1997). bushes and other plants that create resistance and plant growth in the channel is regularly removed. In environmental terms. but environmentalists argued that if the floodplains had been left in their natural state. whether this is rivers. the bypass channel carries only a sweetening flow and is effectively an artificial wetland. The US Army Corps of Engineers (1995) claimed that the flood protection works (which include levees) on the Upper Mississippi prevented an additional US$19 billion flood losses in the 1993 floods. The views. by 1812 New Orleans was protected from floods on the Mississippi river by 150 kilometres of levees. amenity and recreational grounds. Whatever the method the strategy is to reduce flood risk by changing the probability of property being flooded and to reduce the exposure of property and people to floods. In the past. it may be possible to reconstruct the channel so that the frictional resistance of the banks and bed are reduced. For example. Bypass channels have been constructed to increase the discharge capacity of the river during floods. In other instances.4. the banks were stripped of trees. or reducing the resistance of the channel. and increasingly environmentally friendly methods of maintenance are being adopted (Purseglove 1988). lakes or the sea. the bed has been dredged: again this can cause significant environmental damage. bioengineering are being used for bank protection (Gray and Leiser 1982). the term ‘river improvement’ was generally used to describe such approaches which generally resulted in the river becoming little more than a canal and environmentally dead (Purseglove 1988 ). In the 1970s. dikes or bunds) have been widely used for many years to reduce flood risks in most world regions. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . Such barriers are usually designed to reduce the likelihood of floodwater spreading laterally across floodplains. for example.600 kilometres of the Lower Mississippi River which is the most flood-prone part of the Mississippi River basin. for small watercourses. In particular.5 Separating the people and the threat A common strategy employed to reduce the flood threat is to construct barriers of various types between damageable property and people in flood prone areas and the source of flooding. This option was largely adopted because changes to the main river channel were unacceptable on environmental. but are sometimes in the form of increasing the vertical distance between the ground floors of properties and the floodplain thereby creating a space barrier. the Red River Floodway bypassing Winnepeg is another example. Earthen flood embankments (alternatively termed levees. these options are immensely destructive. then the floods would have been of lesser magnitude and the losses due to occupation of the floodplains would have been low (Changnon 2000). the flood alleviation option eventually chosen for the Maidenhead to Windsor area west of London involves a bypass channel equal in capacity to the natural channel for the River Thames (Gardiner 1994). This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. 4. this was often achieved by building a trapezoidal concrete channel. In normal times. An even more extreme example has been the habit of covering over the river and converting it to an underground culvert (Barton 1984). multi-stage channels which mimic a natural river in their form have begun to be used. these were frequently no more than canals: a concrete lined box designed to carry flows with the least frictional resistance.
A traditional method of protection is the use of rip-rap. Geo-textiles are increasingly being used instead and on smaller watercourses. recognising that eliminating flooding would deprive Bangladesh of a major resource from the agro-ecological point of view (Brammer 2000. Unless the protected area is adequately drained. but the country remains exposed to devastating floods as demonstrated in 1987 and 1988. there will little agricultural benefit from constructing the dike system. and in 1989 the World Bank was invited to co-ordinate the efforts of the international community to alleviate the effects of flooding in Bangladesh. conclusions. the military will be involved and mass-mobilisation of the population is also common. However. This decision recognised the shortcomings of that strategy and the Flood Master Plan incorporates non-structural and well as structural measures. Should the flood go on rising so that it threatens to overtop the dike. If the flood goes on rising. ‘inland flooding’ or ‘drainage congestion’. This is variously called ‘local flooding’. China. A significant problem with polders in particular and dike in general is that they interfere with the natural drainage patterns within the area protected. In this phase. the use of plants to stabilise the banks. In consequence. The first level response is to mobilise response and to carry out an inspection of any dikes are other structures.World Commission on Dams Assessment of Flood Control and Management Options 64 A similar controversy of the long-term effectiveness of embankments exists in Bangladesh where flood embankments have been the cornerstone flood control plans. The aim in this phase is to reduce the risk of a breach occurring. It is also difficult to separate out flood losses as a result of the main river bursting its banks from the crop losses than will occur anyway because of waterlogging. Gray and Leiser 1982). ‘Local protection’ then usually consists of a dike ring protecting a small. whenever a flood occurs a graded flood fighting response is activated. then action will be taken reinforce weak points and to deal with such problems as sand boils. Therefore. Between the 1950s and 1980s there has been enormous government and donor expenditure on constructing embankments along the country’s major and secondary rivers. high value area within a large floodplain. typically on the basis of flood levels: the Netherlands. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . In many countries. After this disaster the Government of Bangladesh accepted the offer from major aid donors of technical assistance to find a lasting solution to the country’s chronic flood hazard. a ‘ring’ dike. Initially the Government of Bangladesh insisted that the country’s rivers must be embanked but the Flood Action Plan which was adopted was not based on the total-embankment proposal. The views. Polders are then areas of land surrounded by dikes. for example. can be used (ARSIA/GEOPLAN/CEHIDRO/WASSER WIRTSCHAFT 1999.379 lives were lost. More than 45 million people were directly affected by the 1988 flood and 2. In some countries. an estimated 133m2 have been used over the last 50 years per linear metre of embankment. an urban area within a largely agricultural area. ‘Summer dikes’ are dikes designed to provide flood protection only against frequent flooding and are usually set some distance in front of dikes which provide a much higher standard of protection. then efforts will be made to raise the crest level of the dike through emergency sandbagging. dike systems remain a primary strategy for flood alleviation and with the aid of flood fighting. many of the dikes on the Mississippi supported a flood exceeding their design capacity (US General Accounting Office 1995). an area that is notionally protected from flooding from the main river may still be flooded either from local precipitation or from minor tributaries flowing through the area. A problem with embankments is that they may be destroyed by erosion as the river dynamically adapts to changing flows. Thompson and Sultana 2000). Flood fighting materials are also stockpiled prior to the flood. bio-engineering methods. in addition to the routine inspections that are undertaken. graded rocks: on one section of the Yangtze. bank protection and river training works frequently have to be undertaken to protect the embanked section of river from erosion. those living behind a dike know which are the sections of the dike for This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. Japan and Hungary all have highly developed systems of this kind.
is that a dike system may pass a flood. an expensive and disruptive activity. as on the Mississippi in 1993. Contingent flood proofing is action taken shortly before a flood arrives: this can involve closing gates. Where an existing system of dikes. There are other types of barrier to flooding. Floodproofing is the act of modifying a structure to protect it from flood damage and to keep floodwater separate from damageable property. In permanent flood proofing. and these may be used to reduce the flood risk along rivers and at the coast (where they are called seawalls). based on the 1993 Mississippi flood. In this comparison. Wolff (1997) outlines a methodology for calculating the probability of a dike breaching and SAST (1994) provides a qualitative assessment. The relevant probabilities of failure include both those of breaching and of over-topping. which exceeds its notional design standard of protection. Figure 19 illustrates this approach for a benefit-cost analysis of part of the dike system on the River Yangtze.World Commission on Dams Assessment of Flood Control and Management Options 65 which they are responsible. One consequence of flood fighting. the building is constructed or re-constructed so as to reduce flood losses. the effects of flood fighting measures must be included. of those locations where dikes are most likely to fail. The views. Flood barriers may comprise concrete floodwalls. walls or barriers is being renovated then the probability of failure of the present system must be compared to that with proposed project. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . overtopping simply being a probability of failure equal to one. This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. conclusions. shutters or barriers across all openings.
World Commission on Dams Assessment of Flood Control and Management Options 66 Figure 17 Comprehensive flood management strategy adopted on the middle and lower Yangtze river: dams. detention basins. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . The views. dikes as well as flood warnings and resettlement This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. conclusions.
g. flood waters may force their way through the ground floor and enter the property that way. such as dwellings. as This is a working paper prepared for the World Commission on Dams as part of its information gathering activities.2 0. 1997. For example.5 0. including by waterproofing walls. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . Floodproofing may be a designed-in feature of new buildings located on floodplains. or they may be retrofitted. US Army Corps of Engineers 1984. gates or other closure devices. depending on the subsoil. Federal Emergency Management Agency 1986. especially when floods are relatively deep. The internal design of buildings may also be altered to reduce the effects of floodwater penetration. The science of floodproofing has been developed to its greatest extent in the United States where various federal agencies provide technical floodproofing manuals (e.4 0.9 0. fitting one-way valves on sewer lines.3 0. for warehouses contingency plans may exist to raise stored goods to a height above expected flood levels. 1998a). Flood proofing. 1998. Contingent flood proofing depend upon a reliable flood warning system.World Commission on Dams Assessment of Flood Control and Management Options 67 Figure 18 Dike failure probabilities for an existing dike system 1 0. or to penetrate those walls. at flood velocities above 2 metres/second and where the product of velocity and depth is greater than 7.7 0.6 0. Laska and Wetmore 2000). and by building boundary walls around the house structure. buildings may be raised on stilts and not infrequently buildings will built on raised mounds or with important areas above likely flood levels: all three are typical of indigenous adaptations to flooding. Further measures may include sump-pumps which begin operating in basements when water levels rise. buildings typically collapse – with masonry or framed structures showing greater resistance than timber framed buildings (Black 1975). may be modified in a variety of ways to reduce the risk of floodwater penetration. Structures.1 0 1000 10 20 50 100 200 500 1 2 5 DIKES DIKES + DETENTION operational Floodproofing is a physical engineering strategy to counter the physical force of floodwater (Sheaffer 1967. However. In extreme cases. 1996. electrical circuits and sockets may be permanently routed and located at high rather than low levels. For example. The views. conclusions. fitting openings with permanent or temporary doors.8 0. Equally. and contingency plans and facilities designed to be operated when a flood is anticipated. Technical advice is required to undertake floodproofing because the hydrostatic pressures created by floodwaters can be sufficient to collapse walls.
federal policies encouraged flood-damaged property acquisition and demolition (so-called ‘buy-out’ policies) (Godschalk et. US Army Corps of Engineers 1998a). precautions must be taken against local scouring around the foundations of columns and the design of any columns must take account of the possible effects of debris deposited by floodwaters. Resettlement. thirty villages have been relocated to higher ground to alleviate them flooding from the Narmadi and Tupti rivers (Tobin and Montz 1997). Finally. Recently a number of automatic devices have been developed for flood proofing individual properties or closing gaps in the flood walls. Whilst in France. the overall approach has been termed ‘floodplain management’. ‘Dry flood proofing’. The scope for such resettlement policies clearly depends both upon the number of people and properties involved and the availability of alternative areas to which they can relocate. such as the top of flood embankments. South Dakota. building standards and other regulatory and planning mechanisms designed to ‘manage’ the use of floodplains. The report by the National Flood Proofing Committee (US Army Corps of Engineers 1998a) contains a table setting out the conditions under which each flood proofing option is a viable option. These included floating doorsteps which rise upward as water enters the property. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . In the Gujarat. in the United States there is an active Association of State Floodplain Managers. or from floodplain to flood-free land is a strategy which is used in extremis. Only the latter two options are feasible where flood velocities are more than negligible and where flood velocities are significant. al. When combined with floodproofing. Again. A related strategy for separating people and property and floodwaters is to seek to regulate and control floodplain land use. and flood warning systems. The North American cities of Rapid City. flood proofing is required for new buildings and substantial modification of old ones in the ‘blue’ zone. Bauwesen and Stadtebau 1996. conclusions. The normal conditions governing resettlement also apply. is not generally recommended for buildings with basements because of the problems of uplift (Bundesministerium fur Ruamordung. thus has the greatest potential in shallow flooding across impermeable soils. In Bangladesh some rural householders are used to dismantling their homes as floods rise and moving them by boat to high ground. The views. In the United States property relocation has become more common. This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. there have been attempts over many years to zone floodplains into ‘red’. Following the Mississippi floods of 1993.World Commission on Dams Assessment of Flood Control and Management Options 68 opposed to raising buildings. the relocation of property and/or property either from high risk to low risk floodplain land. 1999). buildings should be evacuated prior to a flood because of the risk of becoming trapped in the building. In France. normally when frequent and severe flooding occurs. pulling a water proof fabric upward between them. A second Dutch system is a hollow glass-reinforced plastic gate which is normally recessed into the ground but also floats upwards when required. and a condition of the availability of federal flood insurance is that the community adopt floodplain building codes requiring flood proofing. India. Buildings can be raised on extended foundations. stilts or on fill. Nor should light masonry buildings be flood proofed above a height of 90 cms because of the risk of structural failure as a result of the hydrostatic forces. this has not been a feasible option in the USA for constitutional reasons. sealing the walls of buildings so make them impermeable to flood water. Wisconsin and Soldiers Grove Wisconsin have relocated some properties from the floodplain. the attempt has been made to introduce land use development controls. The homes may then be moved back to their original locations after flooding has subsided. ‘blue’ and ‘white’ zones (Torterotot 1993): all development is then prohibited in the ‘red’ zone. and similar resettlement projects have been undertaken in eastern Peninsular Malaysia. Prairie du Chien.
al.g. there has been inadequate research to determine the conditions under which land use controls can be effective in controlling development on floodplains. Useful guidelines on floodplain management have been published by Emergency Management Australia (1999). 1999). For example. There is also evidence of noncompliance with floodplain regulations in the United States (Burby 1998. Godschalk et. where there is internal or external pressure then land use controls frequently do not work very well. It is not necessary to take too seriously the novels of John Grisham or Carl Hiaasen to accept that some degree of corruption in zoning decisions takes place as well. for instance. it has been estimated that 80% of the buildings in Istanbul were constructed illegally. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . (b) developments which may be proposed for the floodplain and (c) the residual risks following implementation of management measures.World Commission on Dams Assessment of Flood Control and Management Options 69 Community involvement and participation in raising awareness of the concepts of flood risk and exposure and in developing ways of reducing the risk is a key component of this approach. assessing the flood risk. there is little point in proposing such a law. and developing strategies to reduce (a) flood hazard potential for existing properties. fertile soils and associated vegetation) and protecting these. given that the majority of informal developments take place in violation of land ownership entitlements. However. Thus. amnesties for illegal developments are fairly routine practice. following the recent earthquake in Turkey. The views. the draft zoning plan for Montpellier showed the university to be within the red zone. Again. Either planning controls are ignored and impossible to enforce or they are corrupted. Building controls usually include setting minimum floor levels. In France. after Hurricane Andrew hit Florida it was estimated that 30% of domestic properties were built in violation of existing building codes (Rutrough 1997). the red zone skirts around the university area (Anon 1993). It would seem likely that land use control is most likely to be effective when it is least needed: it will fail where the development pressures are great. The first is that planning controls are defensive when floodplains are considered. Floodplain management also involves recognising the natural values associated with floodplains (e. Miraculously when the final plan was published. around Lisbon there was a massive invasion of river corridors in the 1960s and. al. there is little reason to suppose that a zoning system will be more effective in preventing the occupation of land. Overall. 1996). Floodplain management involves defining and mapping the floodplain. the outcome is likely to be as shown in Table 15. Floodplain management may be effective but it is by no means always so and much depends upon effective policing to gain compliance with regulations. wetland habitats. There are two separate problems. in Italy. It is easier to pass a law than to enforce it and unless that law can be enforced. There are many examples in the United States and elsewhere where floodplain management has reduced the flood problem (see for example May et. This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. was required under the PER plans. Zoning that prohibited development of a floodplain would be perfectly successful if there is no pressure for development. where a ‘red zone’ where development is prohibited. Again. conclusions. This corruption can be monetary or of influence.
they work best with large scale developments. A second problem is one of scale. However. have been a major problem (Anon 1993). 4. a significant cause of increased runoff is the conversion of the front gardens of houses built around 100 years ago to hardstandings for off street car parking. Again. The views. it is considered that the return period of the event which can be reliably predicted is about twice the length of record. no particular logic for selecting the 100 year return period event for these purposes. On the other hand. as in the case of Montpellier. In England and Wales.World Commission on Dams Assessment of Flood Control and Management Options Table 15 Hypothetical analysis of the likely effectiveness of land use planning (Source: Green and Warner 1999) for Internal desire for development Positive land use controls unlikely to be effective little development to be controlled: land use controls very unlikely to be maintained 70 External pressure development Strong Weak Negative land use controls effective if local population has power little development and effective land use controls Thus. the conversion of back garden lawns to paved patios is also a common practice.g. they are much less likely to work with small scale modifications of existing buildings. Where land use and building controls work. It is not practical to enforce controls over such small scale developments. for instance. the USA) to take the 100 year return period flood as the design standard or for other purposes.6 Economies of scale in flood alleviation This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. Land use control has most to contribute when development is undertaken by companies and then as a way of internalising the costs of flood mitigation to the developer. the chance that a domestic property will experience a fire to which the fire service called out is 1 in 250 per year. The reliability is a function of the length of the streamflow gauging record: as a rule of thumb.4. in Britain. the length of record seldom exceeds 25 years. For example. A third problem is the arbitrary nature of floodplain designations. the 100 year return period is often more severe than the magnitude of the flood which is likely to be possible to predict reliably. Thus. These works may increase the impermeable area of a house plot by perhaps 100%. land use controls are most likely to be effective when the majority of those already living in an area do not see any personal advantage in further development and they can mobilise sufficient political power to enforce their preferences. Other risks that have the potential to cause equal losses to the properties exposed typically involve a lower risk. Thus. however. it may be made a condition of planning consent that the developer either provide localised flood protection or contribute towards the cost of a project covering a wider area. getting agreement of the PER and the compromises involved in drafting the PER. There is. Pottier (1998) analysis suggests that in the areas she studied. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . It is common (e. in France. the effect of PER was to reduce the number of applications to develop in the designated flood risk area by about 50% but not necessarily to reduce the likelihood that planning consent would then be given. in France. in Britain. such as demarcating the floodplain for land use planning purposes. Land use controls appear to be somewhat more effective than this rather pessimistic assessment implies. conclusions.
of course be other reasons than economic efficiency for preferring individual measures to a structural scheme. as they are in the US compared with other parts of the world This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. or flood proofing. up to 10% of the capital cost is required to be spent on flood proofing in the ‘blue’ zone (Torterotot 1993). as shown in Figure 20. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . the total cost of individual flood alleviation measures is a function of the number of properties so protected. In France. ecology and agriculture. the cost of the scheme per property of flood alleviation will fall in the future. The costs of building too are generally lower than on hillsides. where a structural scheme permits the intensified development of the available land behind the defences. Figure 19 Economies of scale in collective versus individual flood response total cost individual flood proofing collective flood alleviation scheme number of properties Landuse planning provisions and insurance schemes may require individual property flood mitigation measures such as flood proofing or individual property raising as a condition of development and participation in the insurance scheme in certain flood risk areas. such as flood proofing. The US Federal Emergency Management Agency estimates that flood proofing adds an average of 5% to construction costs. sewers and roads are lower for flat land as compared to sloping sites. There may. Structural flood mitigation measures can be cheap compared to the additional costs of developing on sloping land especially where the floodplain site already has development and infrastructure is in place (Chatterton et al 1993). There are also potentially significant economies of scale through the provision of structural flood alleviation compared to the adoption of individual flood alleviation measures such as house raising. The cost of channel improvement or embankments is some function of the length of the structural scheme and unrelated to the number of properties protected. Indeed. This may be competitive with the costs of structural flood defence where development densities are very low. for example.World Commission on Dams Assessment of Flood Control and Management Options 71 One reason why development is attracted to floodplains is that the costs of providing infrastructure such as water supply. As Figure 20 shows. the perceived adverse impact of an embankment or channel improvement scheme upon the local landscape. conclusions. Conversely. at a certain level of development. The views. a structural solution will become more efficient than for all property owners to undertake their own flood alleviation measures.
the economies of scale and low cost as a proportion of building construction costs of structural flood alleviation schemes mean that flood proofing is only likely to be a viable option when structural flood alleviation works are not themselves economically justified.222 per property. and also on the ecosystems.. In particular.4. groundwater and surface are to a greater or lesser extent interlinked. on the practical grounds that soil and water problems are so closely interwoven and so pervasive that any single-issue solution may cause more harm than good by unforeseen interactions and impacts . To justify their investment. 4. • both the direct and indirect effects of each option should be assessed. Consequently. a number of principles should be applied: • the catchment as a whole should be considered and not simply the local area where the option is applied. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . changes at one point will have ramifications elsewhere.. or changing the flow regime will have effects on each other. properties are quite routinely raised off the ground on stilts.986. In Lismore.. the neighbourhood option. both This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. Penning-Rowsell et al (1987) examined some local protection options for Maidenhead in the Thames valley.. the processes of soil erosion and runoff. sediment movement and deposition. Finally. the term ‘check dam’ is sometimes applied to surface water reservoirs that are intended to retain flood waters in order to recharge aquifers (Jagawat 1998. The individual protection option requires the construction of some 140 kilometres of walling. river flow regimes and sediment movement are both too tightly linked to be separated. For high density developments involving large numbers of properties. residents had to be taking account of some of the non-monetary costs of flooding. changing either sediment production or runoff production. In addition.. In assessing the effects on the environment. including the deltaic and near shore areas (Table 13). Development on a floodplain may then reduce the likelihood of development on the uplands which may. Updated to 1997 prices. the building of 26 kilometres of bunding. The socio-economic system is also a system. Singh 1998). • the effects on the processes of soil erosion. changing the areas that are developed and the nature of that development. with the consequences of each available alternative option being assessed.7 The environmental effects of the different options “The intention is therefore that this study should be holistic . some 42 kilometres of bunding or wall. and the community option. Penning-Rowsell and Smith (1987) reported that on a strict benefit-cost test in terms of flood losses averted. this approach was uneconomic. and • the assessment needs to be comparative. the costs of flood proofing each individual property were calculated as £15. These figures relate to post-development flood protection and the costs of building flood protection into the initial construction of a development might be expected to be lower. in turn. The impact across the catchment as a whole needs to be considered. need to be considered. Nevertheless. a structural flood defence scheme is likely to be more efficient.. At the same time. Australia. ecosystems develop around the prevailing water regime. as well as on runoff and river flows. Both these two examples are for the costs of retro-fitting and incorporating flood proofing measures into the original construction should be cheaper. the costs of providing bunds around each of the main built up areas was estimated to cost £2. The views.World Commission on Dams Assessment of Flood Control and Management Options 72 such as most of Europe.” (CIGAR 1997).843. Constructing bunds or walls around each neighbourhood was estimated as resulting in an average cost per property of £2. thus. conclusions.
‘Silt arresters’ are micro scale versions of warping dams. Leung (1999) describes the use of check and warping dams on the loess upland watershed of the Yellow River. The area behind the check dams may also be used for farming outside of the flood season but a flood storage reservoir is necessary for full use to be made of this land. both the rates of soil erosion. conclusions. and also increasing the time to concentration. ‘Warping’ dams are small dams (usually 2-5 metres) high across gullies which trap silt so creating fertile new agricultural areas in the form of terraces down the gully (Leung 1999). can have catastrophic consequences in terms of mudflows and slope failure: the 1999 Caracas flood being an example.000 tons/km2. New South Wales 1999). What is desirable now may well also change as we struggle to contain the impacts of climate change on biodiversity. and runoff production.World Commission on Dams Assessment of Flood Control and Management Options 73 increase soil erosion and runoff. and also have the advantage of trapping sediment (Development Alternatives 1999. reducing erosion of the gully walls downstream. Thus. have already been heavily influenced by human activity. Thus. a primary purpose of some forms of dam is to trap sediment. may have little apparent direct effects on the environment. In some cases. whilst dams such as the Aswan High Dam trap sediment and in consequence the Nile delta is no longer accreting. the environmental impacts may be significant (Barbier 1989). not always for the bad. It is not therefore possible to apply the equivalent of George Orwell’s ‘Animal Farm’ principle of deeming all sediment reductions to be bad or to be good. ‘Check’ dams are small barriers which provide a local water resource for decentralised irrigation. the thick loess layers being cut by gullies that may be 200-300 metres deep and annual soil erosion rates can exceed 30. Climate change will affect even those rivers that are currently little affected by human activity and will affect both the hydrological regime and sediment patterns in those rivers. rather than it to being possible to consider some pristine wilderness. Settlement on slopes that currently are held together only by vegetation. mobilisation and erosion river and groundwater flows coast building versus determine limits of salt water erosion. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . The views. the global climate will continue to change until it reaches a new equilibrium state and preserving currently valued ecosystems will pose a major challenge. such as planning controls on the floodplain. it is not. This area is some of the most difficult terrain in the world. trapping sediment or reducing sediment loads is desirable. if the effect is simply to decant development onto upland areas. sediment deposition intrusion into river and on sea floor groundwater Water runoff generation Upland floodplain deltaic area Over much of the world. Even when greenhouse gas emissions are stabilised. or to recharge groundwater. Climate change is likely to have very damaging impacts on the aquatic environment because of the multiple impacts including changes not only in flow and sediment regimes but also on water temperature. in other cases. Table 16 Catchment zones and the importance of flows of water and sediment Sediment sediment generation through soil erosion flow regime patterns of sediment deposition. This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. Such small scale dams also reduce the velocity of the flood waters. the deposition of this material further down the Yellow River causes major problems. whilst institutional changes. Thus. Seeking to meet that challenge may involve attempting to hold the current hydrological and sediment regimes within their current ranges through different forms of intervention. Uplands are frequently ecologically fragile and so too are the slopes unstable.
typically this will have to be in the floodplain and so disbenefits depend upon the environmental value of that site as it is. Detention basins Weirs Artificial wetlands Slowing the flood wave Widening the channel Increased variety in channel form which may be advantageous in some conditions. Disadvantages Reduction in low flows. Potential capture of sediment. the answer to the question of the environmental desirability of the consequences of particular flood management strategies is: it largely depends on local circumstances. the environmental benefit is not proportional to the size of the water body). The views.World Commission on Dams Assessment of Flood Control and Management Options 74 Therefore. Sediment deposition. Generally very damaging. see Acreman et al (1999) for detail. and recommendations contained in the working paper are not to be taken to represent the views of the Commission .g. change in flood regime. scale of the reservoir is important in terms of both advantages and disadvantages (e. if a multi-stage channel then potential environmental benefits. conclusions. Potential capture of some sediment. relatively small water bodies may be important for water birds. reductions in soil erosion and sediment generation. these notes may be highly misleading and in no way replace the need to make a good Environmental Assessment of the comparative impact of the different options being considered. Table 17 The environmental management options advantages and disadvantages of different flood Option Runoff control Reservoirs Environment Advantages Reduction in flood velocities and peak flows. destroying most ecological Depends on the form of the new channel. In particular circumstances. Depends upon area adapted. Deepening the channel Reducing channel resistance This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. Sediment accumulation. Table 17 is however an attempt to draw general conclusions as to the relative effects of the different flood management strategies. Generally very damaging and also unstable in new form so requiring continuing disturbance through dredging. Ecologically very valuable. For all channel changes see Purseglove (1988) and Brookes (1988). Effects of reservoirs on a small scale (see New South Wales 1999).
very valuable ecosystems may have been created in the protected area. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . 1989).5 Enhancing coping capacity The capacity of individuals. However. and appropriate mitigation measures. Resettlement area itself may either involve environmental loss or may result in changes in water and sediment load to river. The campaign aims to raise the awareness of flooding and the agencies that deal with it among the public and to enhance individual’s and communities capacity to help themselves in the event of flooding. in western democracies as well as elsewhere. how to fight floods. Local NGOs with technical expertise can help communities to understand local flooding issues by providing technical assistance and by analysing and interpreting technical information for them as happened in the RIMAC Valley Project (Maskrey. None. organisation including organisation within households. Knowledge can be enhanced through public information programmes provided by national or local agencies or NGOs. within neighbourhood groups and within whole communities as a way of pooling knowledge. individual local people and groups within communities often have local knowledge about local flood This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. Changes form of river bank above and below water particularly where requires bank protection May encourage settlement in an ecologically valuable area. households. during and after a flood event. The UK’s Environment Agency is mounting a publicity campaign involving radio and ‘advertising. posters and leaflets. groups and communities to cope with flooding depends upon their knowledge. multi-stage channels offer scope for environmental gain. including their skills and the physical assets and human support that they can call upon. The views. conclusions. the causes of flooding. Resettlement off the floodplain May allow the re-creation of wetlands and re-creation of natural river. In areas where there is a long history of embankments and drainage. what to do before. those responsible for flood hazard management may be reluctant to share information and decision making about local flood problems and solutions with local people at all or until managers have selected a preferred option for action despite local people’s preference for early. Flood hazard management and emergency response agencies can also enhance local communities’ coping capacities by involving them in decision making on all matters relating to flooding and thus enhancing local people’s understanding of the flood problem as perceived by the managers and the managers’ understanding of how the problem as seen and experienced by local people. However. • Coping capacity can be increased by enhancing any of these elements.World Commission on Dams Assessment of Flood Control and Management Options 75 value. Depends upon the environmental value of the land through which the channel is cut. Bypass channel Embankments/levees Flood proofing Depends on the form of the bypass channel. skills and resources and planning and co-ordinating activities to achieve their optimum use and power in relation to other groups in society. resources. 4. wide and continuing consultation on the local problem (Tunstall et al. the resources they command and their organisation and power: • • knowledge about: how to identify the signs that a flood threatens. 1994).
Establishing in the plan. clear cut roles and responsibilities. The commitment of national or federal governments will also encourage regional and local government and other institutions to take emergency planning and management seriously. police. lines of command and control. A major disaster will involve of a large number of national or federal government departments and require their resources to be mobilised. Current thinking is that floods should be included as part of a generic ‘all hazards’ emergency planning system rather than hazard specific plans being prepared. raising public awareness. leadership and other skills within communities can be enhanced over time by training. 4. response and recovery (Parker. This can be achieved through regular meetings.World Commission on Dams Assessment of Flood Control and Management Options 76 problems and mitigating actions that can be taken and community knowledge can also be enhanced by a process of sharing information. Organisational. Emergency planning and management inevitably involve a wide range of different agencies at regional and local level: local authorities. and post-flood recovery. and communication and co-ordination among these agencies is an important element in emergency planning since responses to emergencies by the responsible authorities have often proved poorly co-ordinated as. should be a regular part of the preparedness phase. training emergency planners and managers. Some NGOs such as OXFAM and the Red Cross are exploring the potential of Community Based Organisations (CBOs) for mitigating the effects of flooding. personal communication). joint training for emergency management staff. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . Strengthening organisation is seen as the key to enhancing the capacity of vulnerable people to cope with flooding by certain commentators (Maskrey. However. Planning needs to take account of this. flood fighting. this strategy will cover flood warnings. A clear commitment by national or federal governments to the emergency planning and management process will enhance to its effectiveness. Emergency planning should be grounded in an understanding of local people’s perceptions and needs in an emergency situation. any necessary evacuation.5. (Maskrey. The experience of the RIMAC Valley project indicates that this process is likely to be a gradual and incremental one which may require the support of an independent NGO.1 Emergency Planning and management A strategy for managing all floods is required. informing people of the location of shelters and evacuation centres and means of evacuation. 1992). This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. Allen. exemplified by criticism in the Bye Report on flooding in England in 1998. medical and other emergency services. educating local people on what they can do to reduce their vulnerability. The views. Therefore. indeed in the RIMAC Valley this was seen as crucial but that this confidence and capacity building can change the relationship between community groups and the state enabling them to negotiate effectively with national government agencies for the resources they needed (Maskrey. Local people are unlikely passively to wait for official advice but will act on their own initiative. rehearsing the emergency response. and rehearsals of emergency planning procedures. Preparedness involves foreseeing and preparing for potential harmful events by drawing up contingency plans. An emergency plan is a prepared social network rather than a document. transport and supplies. as important as an emergency plan is building up understanding and good co-operative relationships between the individuals and agencies that will be involved in the emergency response so that the plan can be implemented in a flexible but co-ordinated way. There may be a tendency for agencies to double count resources available for use in emergencies or to be unaware of resources that could be mobilised. 1989). voluntary organisations and the media. conclusions. and experience focused on the flood hazard and its mitigation. the armed forces. for example. careful drawing up of inventories of resources: personnel. 1989. The development of warning systems is also part of the preparedness phase. Emergency planning and management has three phases: preparedness. 1989).
relocation out of the flood risk area may be part of the recovery process. The views. or prefabs may be an important element in the recovery. the provision of emergency shelter and emergency feeding arrangements and /or evacuation (Drabek. and avoiding health and stress effects when flooding threatens. the repair of flood damaged and renewal of contents. search and rescue activities. the dissemination of warning. greater emphasis has been placed in flood hazard management on the improvement of the technical flood forecasting element. Even in the relatively moderate and short lived flood events that occur in the UK. flood warnings have to be delivered.5. to the floodplain occupants affected: businesses. The recovery phase is likely to last the longest. this only becomes a ‘flood warning’ when it is received by those who need it in a form which they can use. A flood warning system should be a viewed as ‘total system’ (Emergency Management Australia.and off-site emergency plans are essential for dams (Division of Disaster Emergency Services 1986. it can take households more than six months to get back to normal. and includes a process of review and improvement. 1986). In practice. and. the problem is usually the opposite one: of persuading the public that there is a potential hazard. The public who are at risk must be made aware of how they will be warned and what they should do if a warning is necessary. Keys. communications and medical services will be other priority activities. On. service providers and residents. However the dissemination phase has frequently been found to be the weak link in the flood warning chain. where possible. Federal Emergency Management Agency 1985). Handmer. Where land is available and local people are willing to move. transport. The restoration of services such as power. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . In addition. rather than on the complex social processes involved in disseminating warnings at regional and local level (CNS 1991. Until recently. institutions. Where houses have been destroyed or need long term repairs. Flood warnings are issued with the overall objective of saving life and property losses. The earliest reported system of flood warning was a system of horse riders stationed along the Yellow River to ride downstream. the provision of temporary accommodation in hostels. The recovery phase will involve the clean up operations. 1997).2 Flood Forecasting and Warning A ‘flood forecast’ is a prediction of future flooding. There is frequently a fear by officialdom that the public will ‘panic’ if they are told that there is a risk. For example: This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. either directly or indirectly.World Commission on Dams Assessment of Flood Control and Management Options 77 Examples of the kind of activities likely to be undertaken during the response phase include the mobilisation of emergency personnel and resources. water supply and sanitary services. 1997. food supplies. the response of agencies and the public in the threatened area. conclusions. tents. 1995) which integrates: • • • • flood prediction and the assessment of likely flood effects. Specific objectives are to elicit appropriate responses from the people and organisations involves. Flood warnings have to be disseminated both within organisations and between agencies with different responsibilities in flood hazard management and response. 4. usually undertaken by a national weather forecasting body or other national agency. caravans.
believe the message and believe that there is a personal threat to their household. • • Flood warning systems are usually and most easily developed to cater for areas where flooding is relatively frequent where it can be argued they are less needed because the frequently flooded will know the signs of and what to do in a flood. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . farm. conclusions. The views. as the trigger for flood fighting activities by individuals and organisations such as dike protection. to initiate flood proofing activities on individual properties. people must: • • • receive a warning message in time to act. Pd is the probability that the forecast is disseminated. An initial requirement for effective flood warning is that the areas and properties at risk should be identified by the flood warning agencies through mapping. 1986). to trigger compulsory or advisory evacuation out of the flood risk area. research evidence clearly shows that people normally do not panic on receipt of a warning but denial. and/or to stimulate property saving activities such as moving livestock. to alert people at risk to the need to listen for and seek out further information and advice on the emergency. to encourage organisations and individuals to alert others at risk to the danger. continuing as normal and seeking confirmation are more common initial reactions. including the local authorities.World Commission on Dams Assessment of Flood Control and Management Options • • • • • • • • 78 to alert operational staff to undertake operational tasks such as opening weirs and operating detention basins. CNS (1991) presented a simple linear model for estimating the likely benefits of flood warnings in terms of likely damage reducing action taken: Pf * Pd * Pi * Pa * Pc where: Pf is the probability that an accurate forecast is made. emergency services. this requires the technical forecast to be interpreted in language that can be understood by laypersons. It is less common for flood warning to be successfully provided for more extreme events or for defended areas where the need for warning may be greater. however a need to have a warning strategy for such events. modelling and past records. to trigger removal to a safe shelter within the flood risk area. business stock and household effects to a safe place. multiple but consistent warning messages. understand the warning message. clothing and essential medicines. and messages from a credible source which contain specific local information are more likely to be believed (Drabek. water. For individuals and organisations to decide to act appropriately on receipt of a warning requires the following assuming the flooding has been forecast successfully. and be physically or mentally able to take the appropriate actions. There is. This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. business or organisation. A further requirement for effective flood warning and response is that the roles and responsibilities of the flood hazard management and emergency response organisations. know what is the appropriate response in the circumstances. together with emergency supplies such as food. Dissemination systems need therefore to provide means of confirmation. through previous experience or public awareness raising campaigns. armed forces. should be planned for and each must understood their role.
Of those aged over 85. more than 70% of both men and women have a disability. and recommendations contained in the working paper are not to be taken to represent the views of the Commission .World Commission on Dams Assessment of Flood Control and Management Options 79 Pi is the probability that a member of the individual household will be available to be warned. This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. conclusions. a warning issued when there is no one there to receive it will be ineffective and it is difficult to give an effective warning during the sleeping hours. Together. In addition. The views. Pa is the probability that the individual household is physically able to respond to the warning. in the UK. 18% of the population have been assessed as having a moderate or serious disability.g. with 4% of men and 5% of women having a serious disability (Department of Health 1997). Anderson et al 1994). and Pc is the probability that the individual knows how to respond effectively. some 1 in 6 people in the USA suffer from arthritis or rheumatism (US Department of Health and Human Services 1999). the PRA and PHR coefficients set the upper limit on the potential effectiveness of a flood warning system but only the PHR coefficient is fixed. For instance. The PHR coefficient can be approximated from national statistics on disability and chronic ill health. the likelihood of a person suffering a disability is related to their socio-economic conditions. Household time budgets can be used to estimate the PRA coefficient (e.
the establishing a good understanding between forecasting and warning agencies and the national ad local media is important. conclusions. Sorensen (1988) has estimated. The reliability of the flood warning dissemination system has been shown to be low and variable in UK research with very few cases in which an official warning reached a majority of those at risk (Penning-Rowsell et al. Local and national broadcast media are an increasingly important and credible means of flood warning dissemination.29 of the potential for residential properties (Parker 1999). the proportions of the population who had evacuated an area against the elapse of time. However. This is because This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. 2000). for example. Where warning messages are passed indirectly to the public it may be necessary to build in redundancy through multiple message passing channels both to reduce the likelihood of failure and reinforce the message. The coefficients vary with the lead time available (Figure 21). it is unlikely that on average residential damage savings from flood warnings will increase much by increasing warning lead times beyond four hours. In the UK availability of people to receive and respond to warnings. Therefore. Damage reducing activity is likely to build up over time as householders confirm and begin to take the warning seriously. Indirect systems are likely to take longer and links in the system may fail. the same result has been found in studies in France (Torterotot 1992). and Smith and Handmer (1986) have estimated the likely effectiveness of flood warnings as function of previous flood experience and warning lead time. The views.. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . Direct warning systems are likely to be more effective than indirect ones which pass through intermediaries.World Commission on Dams Assessment of Flood Control and Management Options Figure 20 Probability of successful response as a function of warning lead time 80 Probability Time The CNS model has been calibrated for the UK on the basis of empirical research on flood events in England and Wales affecting residential and commercial property although less evidence is available on non-residential property and non-residential estimates must therefore be treated with caution. significant proportions of elderly and incapacitated households resident in flood risk areas and failure to respond effectively severely limit the damages avoided to 0.
Again. 1977). Parker and Handmer conclude that rather than working in competition. digital television will enable CEEFAX to carry detailed information and the expansion of access to the internet also offers a way to provide detailed information that can be accessed by the public on demand. such as a siren or a radio message. Where beliefs and expectations are not appropriate then action will need to be taken before the flood to change them. The warning process has also to be understood as a two-way communication process rather than one of simply disseminating a message telling the public what to do. The views. Both the public and the institutions may believe that the dikes have reduced the risk of flooding to zero. a more extreme flood will require that they evacuate instead. Research reviewed by Parker and Handmer (1998) shows that unofficial warnings systems incorporating locally relevant and appropriate knowledge and extensive personal communication networks are widespread in many different physical and cultural settings as well as in the UK. than a system that can deal with the requests from the many for more information. If their expectations are appropriate then the flood warning can simply reinforce that expectation and stimulate action. and tiredness and lack of secure storage space may put a limit on further damage reducing action except where outside assistance is received (Parker. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . It is easier to design a flood warning system that disseminates a single message to many people. Figure 22 illustrates this process for rainfall initiated events. A second problem is where they have some level of flood mitigation but then a more extreme flood occurs. A flood warning system should not be based on this illusion. and to get further information. the evidence suggests that it is more fruitful to find ways of combining official and unofficial systems to take advantage of the strengths of both. Those disseminating warnings have to trade off the length of warning lead time given against the certainty of flooding. particularly relatives. It also needs to be based upon an understanding of people’s beliefs about flooding and their expectations of the most appropriate action to take in a flood. similar diagrams can be prepared for snowmelt and forms of flooding. For instance. in particular. People seek to cope with floods and will not follow official advice and instructions unless these are regarded as credible. when there is a risk of a dike breaching. 1998). increasing the lead time and hence increasing the potential effectiveness of the warning can only be bought at the cost of increasing the uncertainty of the prediction. Floods also set off attempts by the public to both to communicate with others. Conversely. or serving as alternatives. people sought to outrun the flood rather than escape it by climbing the side of the valley.World Commission on Dams Assessment of Flood Control and Management Options 81 householders are likely to have moved all the easily moveable items within four hours. In the future. people may have learnt that they can sit out a flood by retreating upstairs or to the roof. 1991). it will often be necessary to either encourage people to evacuate or to evacuate them. conclusions. there can be circumstances where people’s beliefs are wrong and their expectations are wrong: for example. a major problem in 1998 Easter floods in England was that these were extreme floods affecting areas for which no flood warning system had been developed (Bye and Horner. A not unusual response to receipt of a warning is therefore an attempt by the recipient to find what it means. The available lead time between the earliest possible precursor event that can This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. Two particular problems are when the public (and institutions) have experience of floods and have adapted to that risk but then a flood occurs to which an entirely different form of adaptation is required. in the Big Thompson flood (Gruntfest. This requires a willingness on the art of warning authorities to analyse and design flood warning systems within the local community context blending ‘top down’ with ‘bottom up’ method. A pervasive institutional myth is that in a flood that nobody will do anything until they are told what to do and then they will do what they are told. Potential advantages of developing flood warning systems with the involvement of local people are that this will ensure that warning messages are understandable and appropriate and that dissemination methods meet the needs of local people.
drinking water and sewerage services. through loss of power. and where other natural or artificial structures fail. they leave the area. in China where large-scale evacuations are carried out by the army Whether upwards evacuation is a safe alternative depends upon the nature of the flood and the form of construction of the buildings. cyclone shelters on the east coast of India and in Bangladesh serve the same purpose and are provided in China as a place of refuge from river flooding. in the Netherlands. the population could seek refuge in these areas. Evacuation may be precautionary executed prior to the onset of a life or property threatening event. Summer 1997). for example. or take place during the flood as rescues. or as aftermath evacuations undertaken when living conditions in the affected area prove intolerable. Thus. Evacuations may be voluntary or compulsory as for example.World Commission on Dams Assessment of Flood Control and Management Options 82 be used to predict a flood and the arrival of that flood varies from catchment to catchment. New technology and advances in knowledge are yielding some improvements (Larson 1997. in the case of dry weather failure of dams. continues to be very difficult because there is so little time between rainfall and a flood (Landerarbeitsgemeinschaft Wasser 1995. steep flashy catchments. It has the great advantage that it takes much less time to execute than outwards evacuation where evacuating several hundred thousand people may take several days. warnings are most likely to be effective on lowland rivers and they are most problematic on: flashy rivers. in small. weather radar predicted rainfall recorded rainfall upstream flows local flows elapsed itime increasing certainty 4. Figure 21 Data required in order to increase warning lead time telemetered rain gauges telemetered stream flow gauges direct observation satellite data. Flood embankments themselves are This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. Obled and Tourasse 1994). a traditional adaptation to the flood risk was the construction of earth mounds. a flood refuge mound. Providing flood warnings in one of the circumstances when they are most needed. shelter and transport mobility. The views. if the dikes were in danger of failing. conclusions.3 Evacuation Evacuation can be upward or outward.5. In Zeeland. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . The success of flood warning systems where there is less than four hours between the precursor event and the onset of flooding is problematic. individuals may shelter in a safe place within the hazard area: a higher floor. or a typhoon shelter in Bangladesh where they may be trapped until the flood waters recede. Alternatively. Today.
1986). socially stressful and carry with it the possibility of disease spreading (Fordham and Ketteridge. Bellamy and Harrison (1988) have estimated on the basis of past evacuations. evacuate to stay with family or friends (Drabek 1986. Sorenson‘s studies of US evacuations show the evacuation rates achieved to vary from over 90% to less than 30% within two hours. Families frequently have to split up due to lack of emergency accommodation.. Clear advisories from credible agencies which can indicate voluntary or compulsory evacuation also appear to be factors (Drabek.4 Compensation This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. The research shows that. particularly over a long period. pregnant and very young.12 * population requiring evacuation0. Evacuation centre accommodation may be of poor quality. This can cause psychological and social disruption to family life (Fordham 1998. One problem is that people may flee a flood without taking their medication with them. can break up communities’ social networks and support systems and thus reduce communities’ capacity to recover. neighbours and friends in the local area and wider community communicated through the media (Drabek. Evacuation. conclusions. food and a safe source of water supply. In that event. This needs to be taken into account in evacuation planning. Such shelters should be stocked with medicines. overcrowded. Schware. Evacuation itself has a human cost depending upon its circumstances and duration. most typically as family units. Rescues during a flood event may be dangerous in themselves and highly traumatic for those involved (EUROFLOOD.5 Past experience has been found to affect evacuation behaviour but not in a consistent way: those threatened may fail to evacuate because of past experience of lesser floods. The ability of the public to evacuate prior to an event is crucially dependent upon effective and timely warning of the threat. residential and nursing homes and hospitals may not be the most appropriate response in all flood events and for all those affected. access. site characteristics. When people evacuate they commonly do so as group members. 4. or desire to leave one member resident to protect the home or may become separated. Precautionary evacuation. 1999). Drabek (1986) concluded that when warned adequately of approaching natural disasters. Drabek 2000). 1986). disabled.5. the more likely that individual is to evacuate (Drabek. by preference people. Evacuation also disrupts people’s livelihoods particularly in agricultural or fishing communities.000 people and many thousands of cattle when dike bursts and flooding threatened in 1995 in the Netherlands. 1986). mainly self-evacuation of nearly 250. 1986) Key barriers to evacuation are people’s reluctance to abandon their homes. their desire to protect their property particularly livestock and other valuable possessions and fear of looting (Schware. an initial evacuation decision in one area triggered a sequence of evacuations throughout the affected province indicating that the propensity to evacuate is affected by group interactions and the behaviour of others. the rate at which evacuations can be executed although such estimates must reflect the physical locations. However. Rosenthal et al (1998) concluded that householders’ and authorities’ experience of a serious flood in 1993 contributed to effective planning and the smooth. Tapsell et al. 1982. 1996). and recommendations contained in the working paper are not to be taken to represent the views of the Commission . The research literature shows that families may delay evacuation in order to evacuate as family units or to account for missing members before leaving (Drabek. transport available and the nature of the threat investigated: Evacuation rate = 14. The literature shows clearly that the higher the individual interprets the risk to be. approximately 50% will evacuate upon receipt of official advisories.World Commission on Dams Assessment of Flood Control and Management Options 83 similarly used in Bangladesh. Evacuation rates in flood events and other natural disasters vary greatly and multiple factors are involved in explaining why people choose to evacuate or not. although appropriate for life threatening events and for vulnerable groups such as the elderly. The views. 1996). 1982).
Fund organisations do not necessarily have the expertise of insurance loss adjusters in assessing and processing claims. conclusions. The extent to which government is decentralised and lacking powers over landuse planning which may lead it to seek to link an insurance scheme to floodplain development control. it is expected that governments will compensate or aid the victims of disasters. Thus. The capacity of a country’s insurance industry in relation to potential. which were often set up as successors to such funds (5. with the notable exceptions of Australia and the United Kingdom. The relationship between the government and insurance industry.World Commission on Dams Assessment of Flood Control and Management Options 84 In many countries. To this aid may be added compensation in cash or kind via international organisations such as UN agencies. The level of flood risk and thence the degree to which the government becomes involved. flooding and flood defence are characterised and traditions of ‘communal’ or ‘national’ solidarity or private responsibility (Green. the simplest model of society may be seen as a mutual protection society. corruption charges and equity issues concerning payments are common problem with relief funds (Fordham and Ketteridge. whose members come to each other’s help in the event of a disaster. In many flood disasters in both developing and developed countries. The views. 4. 1996 ). how floodplain development. 1999). Ideologies which shape the conception of floodplain problems and the response. Such problems lead the Belgian Government to consider an insurance based scheme like the French scheme after floods in 1995 (van Hassel and van Lindt 1998). through national disaster relief schemes which have been set up in many countries or through local disaster relief schemes set up by local government agencies or voluntary relief funds.2. Belgium similarly has a National Disaster Compensation Fund to which victims can submit claims once a disaster has been declared. not least because such compensation is generally only available when a ‘disaster’ is officially declared.2. In contrast.1). The resources and capacity of national government in relation to the potential for losses. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . The slowness of compensation processes. particularly catastrophic losses and whether or not the country has access to global insurance markets. has never accepted a responsibility for compensating its citizens affected by disasters. for example. the Netherlands’ Constitutional Court has held that flood insurance is illegal because the State has a duty to protect its citizens and to throw the burden of providing for compensation for losses onto the citizens would be to attempt to avoid this duty and victims are compensated through a National Disaster Fund. There are a This is a working paper prepared for the World Commission on Dams as part of its information gathering activities.5 Flood Insurance The insurance industry is in business to make profits and not to take risks unless doing so will be profitable. An alternative to compensation through relief funds is provided by various forms of insurance schemes.5. In general. international NGOs other national governments and NGOs. little or no compensation will be available to those affected. The UK government unlike the governments of most other countries. the insurance industry has considered flooding to be an uninsurable risk except where the government controls the level of risk to which the industry would be exposed. The role that compensation and flood insurance plays in flood hazard management varies from country to country depending upon a number of interrelated factors: • • • • • • • Whether or not the Government has a duty or a tradition of providing disaster relief. organisational and financial problems.
as in Italy. the risk to the individual property is independent of that of other properties and the individual loss is small relative to the total income from all insured properties. the insurance industry is specialised in processing and paying out claims whilst governments are not. The US National Flood Insurance Program is best seen as a solution to a particular and unique US problem. in Hungary it has been possible to buy cover against the risk of being flooded as a result of a failure of the dike system. although the US government now carries more insurance exposure – one estimate being that its exposure now amounts to $US 520 billion (Wright 2000) . such as oil rigs or satellites. Therefore. in practice few governments could leave claims unpaid because of the bankruptcy of the retail insurer. with risks such as fire or burglary. Firstly. these are generally the poorer parts of society. Spain and the USA. The US programme is not strictly an insurance programme. in contracting out the distribution of compensation to the insurance industry or to other financial institutions. then it is suspected that few people could afford to pay them: the risk of being flooded even on a protected floodplain is generally several times higher than of experiencing a major fire. In France. such policies were offered for a period but the company withdrew after the Meuse floods. Constitutionally. since this is in effect a tax paid to a private company.than the capacity of the global insurance industry. The views. there are two advantages. total annual losses are not subject to wide swings. it may be preferable to shift this cost from the taxpayers on to the purchasers of insurance. those in the same catchment. is a cost borne by Government. Generally. where flood insurance is available it is through some form of public-private partnership and insurance premiums are generally subsidised either by the tax payer or across the wider premium pool. conclusions. the insurance industry is willing to provide cover against the residual risk. Secondly. The problem will remain about what to do about those who are not insured. But because the risk to one individual property is not independent of the risk to other properties. It is not possible to require everyone to purchase insurance. necessary if the risk is to be identified. In the Netherlands. as in France. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . the information with which to make a detailed risk assessment is generally not there anyway. In the case of large single risks. Thus. The risk accepted can also be offset by reinsurance. except to cover for the potential losses they impose on other people (e. Secondly. Where governments have accepted a duty to compensate their citizens for losses in disasters. Spain and the UK. flood cover is subsidised by through an additional loading on all purchasers of insurance. if actuarially fair premiums were to be set. floodplain mapping. This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. Secondly. the theory behind the insurance programme in the USA is that the Federal subsidies to the insurance programme will be less than the resultant savings to the Federal government in disaster aid (Czerwinski 1999). but a means by which the Federal government could induce local governments to introduce building controls (but not land use controls) in flood risk areas. For flood risks. where governments have compensated disaster victims. Again. In other instances. land use and building controls are reserved to the States and therefore the Federal government could not require such controls to be introduced.g. Consequently. any form of compensation causes friction amongst the victims and politically it is preferable for a government to be able to blame problems on another party. Thus. Firstly. Thirdly. the premium income from an individual property is relatively small so a detailed risk assessment would not be justified. the individual premium is sufficiently large that the costs of undertaking a more detailed risk assessment and tailored policy can be justified. the French model is probably more generalisable to other countries. as it is not in the UK.World Commission on Dams Assessment of Flood Control and Management Options 85 number of reasons why the industry has generally not been prepared to insure against flood losses. third party car insurance). the government bears some or all of the risk and where the government is not formally the reinsurer. one flood can therefore result in a large loss.
In particular. the Association of British Insurers has indicated that in some circumstances it will refuse to provide cover (ABI 1998). Above this level. conclusions. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . any further coverage would have to be brought on the commercial market (Halcrow et al 1999). The views. it was therefore proposed that Hungary should adopt a tiered approach. Thus.World Commission on Dams Assessment of Flood Control and Management Options 86 Recently. A second layer of compensation could be bought with standard household policies. This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. the developers internalise the cost of flood alleviation. there are some circumstances where it can reinforce an existing policy. A basic level of compensation would be paid by the government to all flood victims. the risk being borne by an additional charge on all household premiums. that it would be prepared to do where development is given planning permission against the advice of the Environment Agency. However. Although generally flood insurance is essentially an ancillary to flood hazard management. the purpose of this threat is less to prevent development of the floodplains than to ensure that in any development that does occur.
international institutions like the World Bank are concerned about the large amount of development-project loans that are diverted to finance disaster recovery. conclusions. Developing and emerging-economy countries are often constrained in raising their budget deficits or borrowing large sums. innovative financing mechanisms to aid these countries in hedging their disaster risks might be considered. The most common way for governments to raise funds after a disaster is to borrow from their central bank reserves or to issue government bonds. Governments can also divert funds from other public programs or impose special disaster taxes after a major flood. This problem is aggravated by the fact that international aid for disaster recovery is generally only a small percentage of the costs. For instance. their main role may be financing disasters that comprise a large proportion of a country’s GDP since after such events the government will have an extremely difficult time raising sufficient funds from its traditional sources. The attractiveness of these instruments will increase if they encourage cost-effective measures for reducing the losses from floods and other disasters. These financing instruments lead to important issues of equity and efficiency. and their use will depend on the availability and comparative cost of traditional financing instruments. the public authorities can borrow funds from international lending agencies. Joanne Linnerooth-Bayer International Institute for Applied Systems Analysis. Especially in poor countries. Given the size of the global capital market (over $US 70 trillion). After a disaster. These instruments are new. There may also be financial and political constraints on ex post. or diverting from the public budget. and have been made possible mainly because of new scientific studies. and taxes are politically difficult to implement. Since the World Bank and other lending organizations are concerned about the losses on their investments in these countries by having funds diverted to disaster relief. such as catastrophe bonds. engineering analyses and advances in information technology. Finally. this will channel funds from international capital markets to aiding the recovery effort. and taxes spread the losses beyond those who are at risk. hedging instruments are costly for governments. Despite their potential. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . Another challenge is presented by the fact that developing and emerging-economy countries will have great difficulties in paying the costs of ex ante transfers of risk. the public authorities often experience difficulties in raising funds to repair damaged infrastructure and to assist the recovery process. In addition. and one of the greatest challenges to financial risk management is linking hedging instruments with loss prevention. borrowing may transfer costs to future generations.World Commission on Dams Assessment of Flood Control and Management Options The financial management of flood disasters: Should governments insure? 87 Throughout the world. but shares the government’s losses by sacrificing interest or principal when catastrophes do occur. the average annual. A catastrophe (CAT) bond is an instrument for transferring catastrophic risk to investors. If a government has insurance or has purchased CAT bonds in advance of the disaster. Alternatively. a government can hedge its risk of incurring large post-disaster expenditures by purchasing traditional insurance or issuing insurance-linked securities. governments incur large expenses following floods and other catastrophes. borrowing domestically or internationally. The views. and there is little private insurance to cover flood damages. financing instruments. The investor receives an above-market return when catastrophes do not occur. Laxenburg This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. governments can mobilize their own financing sources by imposing taxes. global damage from floods (around $US 23 billion) could be easily absorbed using these new financial instruments as sources of funds. There are two principal mechanisms available to governments to fund the costs of recovery: financing instruments and hedging instruments. For this reason. Financing instruments are arrangements whereby the government either sets aside funds prior to a disaster or taps its own funding sources after the event occurs.
6 Matching the Solution to the Problem The three primary principles which should underlie selecting a flood hazard management strategy are: • public involvement. This was found to be the case in research into the Easter flooding in the UK Midland. it has become well established practice to offer counselling to those involved in natural disasters including flooding (National Institute of Mental Health 1979). 1986). Ethnic minority groups may be disadvantaged during the recovery phase by their lack of knowledge of ‘the system’ and by lack of command of the language (Tapsell et al. However. The five key factors are: • the nature of the catchment and climate. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . What strategy is appropriate depends upon a number of factors. Instead. Projects in Nicaragua and Mozambique illustrate the potential of this approach in certain socio-political settings (Maskrey. those living alone or isolated without relatives or friends in the local community to call on for support. In the UK. The lack of practical help with clearing out damaged contents. • whether or not a flood mitigation scheme is already in place. In flood events in developing countries. flood victims are largely left to their own devices to accomplish the clean up and repair and reconstruction of their property and livelihoods in the recovery phase. and • taking a catchment based approach. lack of practical advice on how to dry out and restore their property and to deal with insurance or compensation claims are common criticisms voiced by UK flood victims (Tapsell et al. 1989). The US Federal Emergency Management Agency and the American Red Cross (1998) produces a manual which includes advice on what to do after a flood which might serve as a model for guidance for flood victims in western countries. 1999). UK research has consistently shown that flood victims attach importance to the support they receive from relatives and friends but it has not demonstrated that this support has a significant impact on their overall seriousness of the impact of flooding on their household (Green et al 1994). Those who need most help and support in the aftermath of a flood are likely to be those most vulnerable in normal circumstances: the elderly.6 Help and Counselling In the UK as in other parts of the world. counselling was offered to flood victims in Towyn in 1990 and it is becoming more common to provide such services. take up has often been low because of the perceived stigma associated with using mental health services (Drabek. shelter. but may see help with their most basic needs for food.World Commission on Dams Assessment of Flood Control and Management Options 88 4. 1999). victims may be as or more traumatised. 4. The views. These principles mean that there are no universally applicable solutions. and consequently also the nature of the flood. Strong arguments can be made in favour of mobilising community based organisations to take action in the emergency response and recovery phases of a flood event. Flood victims may feel that others including officials whom they have to deal with do not appreciate what they have been through and may feel isolated as a result. the disabled and sick. This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. and health care as priorities and counselling as less relevant. In the US. what is the appropriate solution will be more or less unique to the catchment in question. conclusions. 1999).5. • appraisal led design. in which some victims expressed a need for counselling to help them cope with their feelings (Tapsell et al. • the nature and intensity of existing uses of the floodplain and of the river itself.
the debris loads deposited by the river. including the most extreme. afforestation will not be an option: normally. Flood warnings will be essential but because of the very short lead time that is possible. The views. the primary option for source control. The narrowness of the valley will also mean that dikes will not be an option. will also need to be considered. the ratio of flood flows to base flows is also typically very high. high flood velocities and significant debris loads including trees and boulders. Bypass channels. only vertical flood walls being possible.World Commission on Dams Assessment of Flood Control and Management Options • • the wider state of development of the country in question. woods. evacuation up the sides of the valley will be necessary and it is unlikely that there will be time to bring in outside resources to help those at risk to leave the area. the uplands of the Lake District in England were forested and it was the introduction of sheep farming that created the present landscape of open moorland. In this instance. in the absence of human activity in the catchment. a flood retention reservoir is likely to be an option to be considered. If the catchment is undeveloped. More probably. it will be less helpful if the cause of flooding is snow melt. In addition. In this instance. and the narrowness of the valley. channel improvements have in some cases proved to be the only viable options but the cost is very high. and associated debris load. mean that partial or complete structural failure of buildings is likely to be a real risk. An emergency plan. This site must be geologically suitable and seismicity. 89 The strategy adopted will generally involve a combination of options rather than. will mean changing the nature of the habitat. forests or shrubs are likely to be the natural habitat to develop. and not just increases the likelihood than a multi-option strategy will be adopted. rocky slopes. development should be encouraged to take place higher up the valley sides although doing so may simply expose the development to the risk from avalanches or landslides instead. reliance on a dam or flood warning system alone. conclusions. If at all possible. the likely reliability of the flood warning system must be expected to be poor. flood warning and evacuation will almost inevitably be required to manage the extreme floods. then the number of people who have to be resettled is likely to be low. Before a reservoir can even be considered. The need to consider how to manage all floods. in flashy catchments. Where the catchment is already developed then afforestation is one option. essentially a fringe of reclaimed land backed by steep. Area A will be affected by flashy floods characterised by very short lead times between forecasted floods and flood onset. but may be from This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. and the (inter)national and other administrative functional and geographical boundaries. Flood proofing buildings in these conditions is unlikely to work because of high flood velocities. both existing and any induced by the reservoir. Figure 13 illustrated a catchment where there are flooding problems in two areas: area A on a tributary draining a steep mountainous area and area B on the lowland floodplain. The absence of forests might imply that part of the catchment is above the tree zone or that tree growth is inhibited by other factors. It is relatively easier to provide additional capacity equal to a few times the annual flood flow than to provide capacity equal to tens of times the annual flood flow. In Hong Kong. If the catchment has little development then afforestation. say. Nor may it be acceptable in landscape terms: for example. channel deepening and widening are very unlikely to be options because of the rocky nature of the substrate. there has to be a site where one could be situated and have a significant effect on the flood hazard at A. in pre-historic times. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . Afforestation here has proved to be highly contentious and in the Lake District it has been allowed primarily in the valley floors rather than on the open hillsides. Detention basins will not be an option since these are essentially the lowland equivalent of flood storage reservoirs. If the area is not developed then ideally this area should not be because floods pose both a significant risk to life and mitigating that risk will be difficult.
World Commission on Dams Assessment of Flood Control and Management Options
indigenous or ethnic minorities. Conversely, because the catchment is undeveloped, the likelihood is greater that areas of critical natural capital would be threatened by the reservoir. On the other hand, if the catchment is developed, then more people will be affected by a reservoir but it is less likely that are any remaining areas of critical natural capital. Table 18 options development Appropriateness of alternative flood management options undeveloped floodplain developed floodplain consider whether development consider managed retreat might be better attracted elsewhere; recognise value of any existing wetlands vulnerability n/a consider whether vulnerability is primarily due to flooding or whether it is a wider problem to tackle which other strategies might be more appropriate i.e. farmers may be ill-placed to survive crop failure from floods, droughts, pests. source control will have to take place higher up the will have to take place higher up the catchment catchment storage contains natural storage which will look at flood detention basins; more probably be reduced by feasible in developed economies development characterised by large farms and low rural densities in than less developed countries where the opposite is generally true. Where major effect from upstream areas or tributaries, consider reservoirs slow flood wave n/a may conflict with navigation and other uses embankments if it is appropriate to convert conventional solution wetlands to other uses /walls land use control attractors rather than constraints too late except at the margin may be a more effective approach resettlement n/a may be viable at the margin in urban areas; in rural areas in developing economies, the numbers involved are likely to be large and there is likely to be little other arable land available on which to relocate the displaced population house raising, flood likely to be appropriate if consider localised protection proofing development will be limited; around areas of high value (e.g. development may be constructed on urbanised areas, major industry) artificial mounds etc flood warning essential complement to other strategies There are likely to be more feasible options in a lowland area than higher up the catchment, not least because intervention is possible, in principle, anywhere higher up the catchment from the area under study. There is also more time before the flood crest arrives which makes it more likely that flood
This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. The
views, conclusions, and recommendations contained in the working paper are not to be taken to represent the views of the Commission
World Commission on Dams Assessment of Flood Control and Management Options
warning and evacuation may work. Table 18 summarises the options that may be possible in regard to area B. Because the appropriate option, or frequently a combination of options, should emerge from the examination of the particular conditions, it is impossible to give a comprehensive breakdown of conditions and the options that are likely to be appropriate. The basic rule is to make a preliminary assessment of as many options as possible, and to be imaginative in developing new options; thus, Gardiner (1994) reports that up to 400 options were identified before one option was selected for the Maidenhead-Eton reach of the Thames. The following listing of the conditions when specific flood management options are most likely to be appropriate should be understood in the light of this general principle rather than being interpreted as being prescriptive or exclusionary. It must again be stressed that decisions involve a comparison of available options and a choice between these options; no option should ever be considered in isolation. Source control may be appropriate when: • the quantity and rate of runoff has been affected by human (or other) change over a significant proportion of the catchment; or • there are soil erosion problems (e.g. in steep, unprotected and unstable valleys); or • there are also problems of agricultural or urban pollution; or • where runoff is too valuable a resource to allow to be lost to the sea. Dams may be appropriate when some of the following conditions are present: • a major part of runoff comes from a steep catchment; • the time to concentration is short; • multiple tributaries contribute to the local flood problem and it is important to avoid the flood crests from the different tributaries being synchronised; • the ratio of flow in an extreme flood to the flow of the annual flood is high; and • when the floodplain is heavily developed. Detention basins may be appropriate when: • a major part of runoff is generated in lowland area; • there exist parts of the floodplain which are in a relatively low value use such as large scale farming; • other benefits, in terms of biodiversity or recreation, could be created through the construction of an engineered wetland; or • multiple tributaries contribute to the local flood problem. Channel modifications may be appropriate when: • the watercourse concerned is relatively small; • when the river has already been heavily modified in the past; and • when a bypass channel could be created; or • where a multi-stage channel could be created, or the river otherwise ‘restored’. A practical problem with channel modification is that administrative boundaries typically lie down the centre of the channel and these will either need to be changed or a super-ordinate administration introduced.
Dikes and flood embankments may be appropriate when: • arable land is scarce (in which case the use of ‘summer dikes’ and a retreated main dike line should be considered, as is the practice in Germany and the Netherlands);
This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. The
views, conclusions, and recommendations contained in the working paper are not to be taken to represent the views of the Commission
World Commission on Dams Assessment of Flood Control and Management Options • • •
on lowland rivers; the floodplain is already heavily modified and intensely occupied for essentially urban uses; and on large rivers or rivers where, because of previous development, the river corridor is constricted;
In developed countries, the use of dikes primarily to protect agricultural land can seldom be justified beyond the standard of protection offered by ‘summer dikes’ (e.g. 5-10 year design standard of protection); land drainage of poldered areas is essential for this strategy to yield any significant benefits. Housing raising may be appropriate when: • there are a few isolated and scattered properties on a lowland floodplain. Flood proofing may be appropriate when: • the properties at risk are historic buildings in urban areas, and where these are closely related to the river; or • there is a low density of development; and • buildings are of masonry construction; and • built on a suitable sub-soil; and • flooding will be shallow; and • flood velocities will be low. Localised protection may be appropriate where there are: • isolated high value communities scattered across a floodplain. Managed retreat and possible resettlement may be appropriate where: • the area is one of low population density, generally in rural areas (e.g. the country is highly urbanised); or • developments are temporary (e.g. camp sites, mobile home parks); or • where buildings are scattered or of low value; or • there is a very high risk to life (e.g. in small, flashy catchments), particularly where the likelihood of successful flood warning is low; and • managed retreat would allow the creation of an important environmental resource such as a wetland; and • it is possible and economically viable to provide localised protection to isolated high value developments. Flood warnings are appropriate in the following conditions: • always. They are essential to reduce the risk to life when: • buildings, particularly dwellings, will not provide a safe place of refuge in the event of a flood; • especially when existing properties are single storey (including tents and mobile homes) or of non-masonry construction; • flood velocities will be high and sufficient to cause the partial or complete structural failure of some buildings; • there are transport links across the flood risk area; • on small flashy catchments (including arroyos and wadis in arid regions); • in urban areas where significant activity takes place below grade (e.g. underground car parks; metro systems, shopping centres etc); • below dams;
This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. The
views, conclusions, and recommendations contained in the working paper are not to be taken to represent the views of the Commission
• below artificial or natural dams.World Commission on Dams Assessment of Flood Control and Management Options • • behind dike lines or in poldered areas. in highly developed countries a publicindustry partnership may help transfer flood alleviation costs to developers since if the insurance industry refuses insurance cover then mortgages or construction loans will not be available. and • around hazardous facilities. and particularly. As an illustration of the resulting differences that are likely to be appropriate. The views. • they should be based on a holistic planning process and not solely based upon a consideration of the flood risk. 93 Evacuation (either horizontally or to flood refuges) should be planned for where: • all areas where buildings will not provide a safe place of refuge in the event of a flood. and • in areas of high development pressure. It has been stressed that the appropriate flood management strategy will be dependent on local conditions. for extreme floods outside those previously experienced. the use of growth attractors may be more effective than applying planning constraints. This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. However. conclusions. Table 19 draws some general conclusions as to the strategic management responses that are likely to be appropriate in four different floodplains. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . Where land use controls are considered then: • their primary effect is likely to be transfer some of the costs of flood alleviation to the developers. and • behind dike lines or natural crest lines and in depressions. and in areas frequently flooded. Land use controls may be appropriate where: • development is undertaken primarily by developers rather than by individuals (and it is pointless to mandate land controls where development is primarily informal). Flood insurance is: • largely irrelevant to flood hazard management.
579 km2). China**** Maximum recorded flow = 110. conclusions. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . river. 90120 days). 70. Thames and Yangtze Mississippi valley. Average fall 0. seasons: snow melt in spring.000 m3/sec (1870).000 m3/sec (1927). The views. 2 2.950 km2. Rainfall: intense thunderstorms Drainage area (at upstream Average fall 1. 52% of Hungary is at risk of flooding. so resulting in high Lowland catchment Ukraine and dike on runoff. 4. narrow bypassing dike system on the river valley. international boundary seeks to which is almost exclusively put avoid flooding in Ukraine to arable or urban use. Drainage area = 1.2m/km. Flooding ‘imported’ from the saturated. Local Urbanised country with very Poor rural area with dike system Densely populated (1000 Densely populated area in This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. Comparative flood problems on the Mississippi.900.360 m3/sec on entering 718m3/sec (1947).000 km drainage area country. July-August country border = 9. Emergency flood detention basins have been constructed. May and Autumn rains.000 km2 Average fall 2cm/km (altitude at Three Gorges 40 metres above sea level. at Greatest floods have generally downstream country border = resulted from rainfall on ground that was either frozen or already 138.000. very wide valley and very low river gradient. Ice floods have been a major problem in the past. USA* Upper Tisza valley. Flooding can last 50-120 days in lower part of the Tisza.62ft/mile. Tisza. Thames valley.707 km2. Diked since the end of the last century.g. long duration floods (e. 1600 km from sea) Subject to intense rainfall during cyclone season.World Commission on Dams Thematic Study of Flooding Table 19 94 Yangtze. Three separate flood Floods may last 6 weeks. leaving country.700 m3/sec on Drainage area = 9. England*** Hungary** Maximum recorded flood flow Maximum recorded flood flows: Maximum recorded flow = Flood problem 3. tributaries significantly effect flood crest.
localised protection of urban areas. Some 30. or deepening or widening the channel were all rejected for environmental. high ratio of arable land per capita and industrial farming. low economic intensity of use (from $65. Further investment in land drainage or flood alleviation is not likely to be a good way of reducing the population’s vulnerability. However. isolated urban communities may be resettled. reversion of much currently protected agricultural land to wetlands or use of ‘summer dikes’ to allow continued arable use. Population at risk in excess of 40 million. From perspective of agricultural productivity it would be more useful to tackle the land drainage problem than to maintain the current standard of protection from the Tisza. national scarcity of arable land. Existing dike system is hundreds of years old in places but offers a very low standard of protection (e. densely populated but still essential rural country (and hence fairly homogeneous distribution of both population and industrial activity). conclusions. Managed retreat.000 people at risk. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . Option adopted involves construction of multistage bypass channel capable of carrying the same flow as the main channel around the town of Maidenhead. The views.8 million in Illinois. 95 people/km2) and relatively wealthy area in a highly urbanised country. Policy of raising and strengthening the This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. Currently.g.000/km2 in North Dakota to $1. against 10-20 year return period flood). expected accession to European Union will make current agriculture uneconomic. The river is important in both amenity and cultural terms. providing a high standard of protection (100 year return period) but major land drainage problem. small. China’s flood management policy involves a combination of most available options. Both a history and a potential for catastrophic loss of life because of the depth of flooding (4-7 metres) to be expected if dikes breach. heavily managed river with channel modifications and dike systems. and heavily modified. a problem compounded by the density of settlements. Multiple planning constraints in the form of designations of areas as of outstanding natural beauty.World Commission on Dams Thematic Study of Flooding conditions low population density particularly in area at risk (varying from 3 people/km2 in North Dakota to 78 in Illinois). the depth of flooding expected plus the flood velocity would probably make localised protection uneconomic. amenity or other grounds. plus both flood storage reservoirs and detention basins. localised protection and shift to other forms of investment. Managed retreat would either involve both massive resettlement or localised flood dikes. Probable solution Managed retreat. The floodplain is also heavily used with extensive areas of past and present aggregate quarries. archaeological. both as an historic navigation and as an integral part of the water resource system. as ‘Green Belt’ and for other reasons tended to direct development on to the floodplain Dike raising. which might include development of ecotourism.
US Army Corps of Engineers 1995.World Commission on Dams Thematic Study of Flooding 96 existing dike line may be the best way of buying time until population pressures ease and per capita income increases substantially. The views. Halcrows/FHRC 1996. Shaw 1998. sources: Gardiner 1994. Vituki 1998 This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. conclusions. Marsh 1988. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . Halcrow et al 1999. Framji K K and Garg B C 1976.
no option is likely to be free of risk: settlement outside of the floodplains not infrequently involves a different form of risk. the choice between settlement on steep hillsides and floodplains is a debatable one.000 houses a year in 1985 (Rivers Bureau. Again. the designed probability of failure of dams is generally considerably less than that of other options. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . there are four major potential drawbacks to be considered: • resettlement. although other means of rainfall harvesting (Bruins et al 1986. Ferroukhi and Chookalua 1996. an essentially rural economy and a high population density). and/or • the dam can be used for multiple purposes (when annual rainfall is concentrated over a short period of the year. Similarly. Thus. Desai and Ghose 1998. the proposed dike strengthening and raising programme will require 70. The likelihood of one or more dams forming part of the management strategy will be increased when: • the floodplain is under pressure (e. Figure 12 compared the risk to life resulting from the different intervention options.000 people to be resettled (the estimates for resettlement for the Three Gorges project range up to 1. Jagawat 1998) need also to be considered. some 40 million people would have to be resettled (Hong Qingyu and Luo Zhang-Yu 1999). 1985) and. the case for storing available runoff is strengthened. Gilbertson 1986. • morphological impact and groundwater effects and • environmental impacts. These are the maximum theoretically conceivable amounts and have a probability of occurrence very close to zero. where there is a significant population in the risk area below them.World Commission on Dams Thematic Study of Flooding 97 5. there is a scarcity of arable land. Dikes are seldom designed to protect against a more extreme flood than that with a return period of 100 years whilst dams are.g. conclusions. In the case of dams having a flood alleviation role. or • in complex rivers where it is necessary to avoid flood crests from several tributaries coinciding. typically designed to cope with the Probable Maximum Flood (PMF) or Probable Maximum Precipitation (PMP).2 million people). in Japan. in general. This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. The Role of Dams in Flood Hazard Management The previous section identified two contexts in which the flood hazard management strategy might include the use of dams: • in upland catchments. the comparative principle applies: are these drawbacks worse from the option involving dams than from other available options? For example. and the time available to carry out warning and evacuation correspondingly limited (Figure 22). debris flows and landslides were killing 130 people and destroying 1. or • there is a history of intensive activity on and extensive modification by people of the floodplain. if the floodplains of the upper and middle Yangtze were to be permanently evacuated. or there is high variability in annual rainfall. Centre for Science and Environment 1999. The views. Depending upon the local conditions. • the risk and consequences of failure. the number of people who may die if a dam fails may or may not be greater than the number who will die if a dike (or another solution) is adopted instead. when considering the risks of failure. Dams and dikes have the common disadvantage of discontinuity: the transition time from ‘safe’ to ‘failure’ can be very short.
415 m3/sec. it is not unknown for the PMF to be badly misestimated. ‘Dry weather’ failure as a result of foundation or ground failure or of the dam wall itself is a significant mode of failure. a cyclonic rainstorm produced 700mm of rainfall in the catchment in a single day. expected to affect both rainfall and runoff (Arnell et al 1994). The views. overtopping and dam failure caused a 810 metre flood wave to sweep down the valley. In consequence of that change in the flow regime. However. Nuclear engineers have been keen to compare unfavourable the risks and consequences of a dam failure with those from a nuclear power plant.8). will mean that some PMFs have to be revised upwards. as Figure 16 illustrated.2. Here there can be a number of problems. if the flood forecasting system fails to predict the pattern of flooding then the reservoir may be filled too soon so that the flood crest is passed through unattenuated because the storage capacity is exhausted. Reliability engineering approaches to assessing the risk of failure of a dam are increasingly being applied (Safety and Reliability Directorate 1985). Both may then be poorly adapted to a flood when one does occur. Again. killing at least 2. The environmental consequences of the physical interventions are not limited to the point where the This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. A well-known example of such a misestimate was the Warragamba dam. On average. Climate change. however. The operations manual may prescribe slow opening but the operations staff may either not know the procedures or ignore them. the result of proper operation may be a lower but longer flood than would otherwise exist. Conversely. for a dedicated flood storage reservoir. therefore sources are not cited).000 people. the general principle that it is necessary to design for failure is as true for dams as any other flood management option. if a second flood crest is predicted then in order to draw down the reservoir to provide the capacity to store the second flow crest. whilst a dam may be designed to safely pass the PMF. since all are intended to change the natural flow regime in the river and its floodplain in some respect. In general. Firstly. One countermeasure to uncertainties about the PMF is then to incorporate fuse dikes as well as spillways in the design of a dam. The PMF was re-assessed upward to 26. the flow for a given return period event should be less with the dam than without the dam except for the PMF event when there will be no difference in the flow. the nature of dams is such that a substantial quantity of explosive is required before much happens. In theory. All physical interventions for flood hazard management have environmental consequences (Section 4. Thirdly. arguing that the probability of failure of some dams are greater than of the failure of a nuclear power station. However. the Macchu II dam in Gujarat was built with a spillway capacity of 5. Failure under extreme flood conditions is not the only or necessarily primary mode of failure of dams. The effect is to release a wave downstream. both the river and the downstream population will have adapted to the changed flow regime. that the release is controlled can result in a sudden increase in discharge flows.925 m3/sec but before it was rebuilt on this basis. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . some of the stored water will have to be released over a short period of time. Fourthly. The dam was redesigned on the basis of a PMF of 20. Secondly. one of the risks assessed was of a nuclear attack and assessments of the potential implications of terrorist attacks on dams is fairly common practice (the drawback of such studies is that they simultaneously identify the best ways of attacking a dam. the water stored in the reservoir also has to be discharged.World Commission on Dams Thematic Study of Flooding 98 However.5. this only applies to the rising curve of the hydrograph. a gate may be fully opened rather than being opened slowly in stages. A second risk is the change in downstream flow regimes as a consequence of the operation of the dam. Recalculation of the PMF showed that the spillways were drastically under-designed whilst a failure of the dam would result in widespread flooding in Sydney as well as major water shortages.420 m3/sec (Herschy 1998). the flooding downstream of the dam should be less than without the dam. the sediment regime is also changed. for example. which retains the main water resource for Sydney (Sydney Water Board 1985). conclusions. in the case of the Three Gorges Project.
Although the main disadvantage of the institutional interventions is that we do not yet know how to get them to work well. in some cases. they should not be assumed automatically to be environmentally benign. out into the near shore zone itself.World Commission on Dams Thematic Study of Flooding 99 intervention is made but are also felt up. The views. This is a working paper prepared for the World Commission on Dams as part of its information gathering activities.and down-stream of that point. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . they may be but one lesson we should have learnt is that our assumptions usually come back to hurt us. The virtue of institutional interventions is the appearance and perhaps the reality that they offer a free lunch in this respect. conclusions.
resettlement from the unprotected areas of the Yellow river has been under way. Dams on the tributaries of the Yangtze were operated in the 1998 to reduce the coincidence of flood crests. there were 8 separate flood crests. As with afforestation and deforestation.000 km2. In 1997. Therefore. with resettlement of a further 200. Flood Hazard Research Centre. A comprehensive strategy has therefore been developed involving both water and soil control. The Yellow River carries very heavy loads of sediment down from the loess plateaus and the river is consequently very dynamic: over the last 2.World Commission on Dams Thematic Study of Flooding Flood management in China 100 In China. together with their livestock and possessions. A national compulsory tree planting campaign was started in 1981. the Xiaolangdi dam will raise the flood protection standard from 1 in 100 to 1 in 1000 on the middle and lower reaches of the Yellow River. in 1989 the Yangtze Protective Forest Project was approved covering 13 Provinces and an area of 4.000 people. On the Yellow river. 40 state approved basins were established on the Yangtze. with a total storage capacity of 47 billion m3. Similarly. some 80 million people live on the floodplain of the Yellow River. which is in any case scarce. China Institute of Water Resources and Hydropower Research. reducing the flood discharges on the upper middle reaches of the river. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . The flood plains are now centres of population and economic activity. conclusions.500 years. the practical problem has been that what central government may decree is not necessarily actually what happens on the ground. Overall. These include a number of detention basins. the forested and vegetated area had increased by 5%. In other areas. A further 5 year $2 billion reforestation plan was introduced following the 1998 flood. These works include the use of warping and check dams to trap sediment to create fields for cultivation. with 6 million people living in those along the Yangtze. and the detention basins on both the Yangtze and Yellow rivers are themselves heavily populated. along the Yangtze. 100. in 1998. several hundred thousand people had to be evacuated from some of the detention basins as it was expected that it would be necessary to operate them in order to prevent the main dikes from failing.000 m2. In 1998. In places the river is perched 6-7 metres above the surrounding land. few of the main river dikes provide protection against more than 10 year return period flood. breaches in the main dike were limited to that at Jiu Jiang only by mobilising almost the entire population to carry out emergency flood fighting. However. The views. since 1974 a series of flood refuge areas have been constructed to provide places of safety for 400. The Three Gorges Dam will provide storage capacity of 22 billion m3 for the main stem of river. By 1997. In 1998. covering a total area of 12. In 1988.1 million km2 with the target of providing forest cover to 45% of the area by 2050. most cultivable land.000 km2. a complex system of flood management has developed directed through a flood forecasting system. land transfers were frozen and the State Council called for land use plans to be prepared by local governments. it has undergone 26 major shifts in course and the dikes have been breached on 543 occasions. many dike systems having been first constructed hundreds if not thousands of years ago. Along the main rivers. as a precaution. there is a very long history of flood management in China. soil and water conservation works have been undertaken over an area of 186. Zuyu Chen. Traditionally the main storage has been provided by the natural lakes and the artifical detention basins along the river itself. lies in the floodplains of the great rivers and these have been the centres of the development of civilisation. and in parts the height of the dikes is 20 metres. the earliest dating from 1954. Similarly.000 planned over the next 7 years. On the Loess plateau of the Yellow River. In the Yangtze basin. some 1072 flood refuges have been constructed with a total area of 160. with a target of reducing this by 40-50% by 2030 and a programme completion target of the mid-century. Since 1974. flood evacuation routes have been established. The scarcity of arable land has promoted development of the unprotected areas of the floodplains and urbanisation has resulted in expansion of the urban areas over agricultural areas.000 people have been resettled since 1987. Beijing and Colin Green. A major programme of dike raising and strengthening is now under way along the Yangtze. The budget for the first phase is over 10 billion RMB. Middlesex University This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. the capacity of the lakes has been reduced by land reclamation and sediment deposition. together with a reinforced ban on illegal logging which has been a continual problem. the target set for the whole of China was for forest cover of 15-16% of the country by the end of 1999 with the most recent FAO figures giving a figure of 14% for 1995. a reduction of 11% in the sediment load entering the Sanmenxia reservoir has been achieved so far.
An additional implicit principle is that the river should be restored to something nearer its natural form in order to promote ecological regeneration. The plan is based on five principles: • water and land must be managed holistically. However.290 by 2005 3. conclusions. and the European Commission subsequently became a member. Its initial primary purpose was the restoration of the water quality of the Rhine and its tributaries.1 33 530 790 26 67 815 2.500 90 4. Switzerland and Luxembourg. This involves the spending of some $12 billion over a twenty year period to manage flooding in a basin where there is an estimated $1500 billion of assets at risk. The targets to be achieved are summarised below.250 1. • water must be stored in the catchment and along the Rhine for as long as possible.060 100 5 810 540 300 21. following the floods on the Rhine in 1993 and 1995.460 by 2020 11. The views. Dick de Bruin Rijkswaterstaat and World Bank Action River restoration (km) Regulation of use of flood plains Restore flood plains on tributaries (km2) Restore flood plains on main stem (km2) Increase infiltration on agricultural land (km ) Afforestation of farm land (km2) Increase infiltration in rural areas Levies on impermeable areas/creation of compensation areas Increase infiltration in built-up areas including roads (km2) Increase flood storage on tributaries (million m ) Increase flood storage on main stem (million m ) Maintain critical flood dikes and adapt level of protection (km) Flood proofing Prepare flood hazard maps Prepare flood loss maps Improve public awareness Improve flood forecasting and warning service Reduction of peak flood levels: Reduce damages by: 3 3 2 by 2000 1. the Netherlands.016 162 3. and then those on the River Oder in 1997.880 3. and • integrated.880 1.115 5 cms 10% 25-30 cms 60-70cms 25% This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. concerned action across the entire catchment is necessary for success. and recommendations contained in the working paper are not to be taken to represent the views of the Commission .World Commission on Dams Thematic Study of Flooding Rhine Action Plan on Flood Defence 101 The International Commission for the Protection of the River Rhine was formally established in 1963 by the governments of France.490 73 344 1.5 1. • it is necessary to learn to live with the risk of flooding. Germany. • the river corridor should opened out as far as possible to increase live storage in the river. the 12th Conference on Rhine Ministers adopted the ‘Action Plan on Flood Defence in January 1998.
were one option to be so selfevidently superior to all others. • a catchment based approach which takes account of the inter-relations between the proposed project both geographically across the catchment and between the different functional aspects of managing the catchment. However. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . rigorous. the consequences of one or more ‘do something’ techniques are compared to those of some baseline option. the conflict arises varies from decision to decision (Green and Penning-Rowsell 1999). Thus. particularly benefit-cost analysis (Penning-Rowsell and Chatterton 1977. This requires. that the techniques themselves be transparent and also that the analysis be published in sufficient detail that it can be audited by all parties with an interest in the decision. discuss of these points is deferred to the next section. It is generally accepted that pre-conditions for identifying the ‘right’ option to adopt require: • an appraisal led process whereby the multiple societal objectives are used to identify that option which best satisfies those objectives. and • public involvement at all levels of decision making. Economic and other forms of analysis are not then primarily about numbers but about increasing understanding of what the decision involves. as a minimum. have been adopted for a number of other reasons as well. Firstly. project This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. A practical condition of the correct decision being made is that it must be capable of being implemented successfully. because of the level of complexity involved in many decisions so that two primary requirements of a project appraisal technique are that it increases understanding of the nature of the choice that must be made and simplifies that choice to a manageable level. therefore. US Water Resources Council 1983). the use of project appraisal techniques increases accountability both by establishing an audit trail and promoting consistency between decisions. Project appraisal techniques such benefit-cost analysis. is generally seen as the way to determine the right decision to take. Thus. Definition of the appropriate baseline option is therefore critical and depends on the circumstances (MAFF 1999).World Commission on Dams Thematic Study of Flooding 102 6. logical argument is used to select the best of the available means by comparing their desirable and undesirable consequences. Equally. All of the appraisal techniques are comparative. We need to understand what the decision involves before we make it. Secondly. Reasoning. Since there remain major problems in techniques to comply with the second two pre-conditions. and project appraisal techniques. and a recognition of the role of women as a condition of both the process and outcome being just. it is this understanding that is important and not any numbers that may be generated by the analysis. Moreover. Reasoning. conclusions. multi-criteria analysis and environmental assessment are simply different forms of logical argument by which to compare the different available options. the real question is: • how to choose when there is a conflict between the options? We want to make the ‘right’ decision where by ‘right’ we mean choosing and implementing the best of the available options in terms both of the ‘correct’ choice being made and the decision being ‘just’. it has to be seen to be a just or equitable process. the process by which the decision is made must also be right. public involvement in all levels of decision making was determined to be a necessary condition for a just decision process. In Agenda 21. Making the ‘right’ decision? We only have to make a choice when there is a conflict: the available options being mutually exclusive. Why the options are mutually exclusive and why. then we would have no real decision to make. The views. It is because that no one option is agreed to be superior to all others that we have to choose.
the three key dimensions across which consequences will vary are: what. the Net Present Value. But rules have been proposed to ensure a reasonably wide range of options are considered (MAFF 1999. where there is an existing flood alleviation scheme.World Commission on Dams Thematic Study of Flooding 103 appraisal techniques can only identify the best of the options compared. the question of comparing consequences occurring at different times in the future is handled in benefit-cost analysis through discounting. But if our preferences for the distribution of benefits and costs are determined by the shape of the curve rather than the area under it. then discounting gives us a weighted measure of the area under the curve. The views. if a particular discount rate is used. There are an infinite number of such curves which. but there a rather more ad hoc approach is adopted and assumptions being made are often invisible. However. conclusions. The most useful of these is to ensure that decision process is as open as possible and involves the widest possible consultation. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . If we plot the net annual benefits from the project over time (Figure 23). This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. in particular. they vary in terms of what they can encompass and whether the basis for comparison is made explicit or is simply implicit to the individual analysis. for example. will yield the same Net Present Value. on the basis of this curve. recent guidance (MAFF 1999) requires this curve to be plotted. who is affected and when they occur. We have to rely upon the creativity and experience of engineers to invent new and perhaps better options. then discounting at best ignores this concern. The same problem is also faced in MultiCriteria Analysis and Environmental Assessments. they involve comparing consequences which may vary widely in terms of their nature. However. US Water Resources Council 1983). Therefore. even the best appraisal of a range of poor options will not identify a good solution. Essentially. However. the option of managed retreat should normally be considered. Because appraisal techniques are comparative. we may also have preferences for the distribution of benefits and costs over time and this question is not adequately addressed through discounting. who and when. there are no rules which can ensure that the best option is amongst those compared in a given decision. The way in which this done varies between the different techniques of appraisal. One reason for discounting is to determine whether the capital invested in the project would yield a higher return if invested elsewhere in the economy and to this question discounting is an appropriate response. Thus. Any comparison therefore involves reducing quite disparate impacts to some common framework of comparison. a reasoned argument may then be put forward for adopting an option which does not have the highest Net Present Value.
a catchment agency is necessary so that a co-ordinated approach is adopted by the different communities making up the catchment. with one purpose being to support small-scale community irrigation. In New South Wales. This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. conclusions. for such projects lies in redistributional considerations: alleviating poverty. the small dams being re-introduced as part as a form of sustainable development (Development Alternatives 1999) are typically multi-functional. however. If the net benefits of that community and to the nation are considered then there are four possible outcomes (Table 20). It will often not be possible to leave them in an exactly equivalent situation due to. The final case is where there are national gains but local losses. for example.World Commission on Dams Thematic Study of Flooding Figure 22 The distribution of project net benefits over time 104 A weakness of economic analysis is that ignores the distribution consequences of projects. Here. the ‘who’ dimension. Figure 24 shows a local community in which a flood alleviation project is proposed. The logic here is to ensure that a sufficient proportion of the national gains are redistributed to the local community to leave them also better-off as a result of the project. Irrigation schemes necessarily have a negative impact downstream by reducing water availability. The ideal outcome is the ‘win-win’ option where both the community and the nation gain on balance. An outcome where both the community and the nation are losers is obvious undesirable but may occur when other interests have seized control of the decision process. A different problem arises with so-called ‘pork barrel’ projects: those where national resources are used to provide a project which yields net benefits to the community but these are not justified in terms of the national resources expended. The justification. if any. farmers are restricted to impounding 10% of the runoff as a ‘harvestable right’ (New South Wales 1999). a scarcity of arable land. Again. no monetary or other form of compensation may be sufficient for the community to accept this loss. A local community may decide to undertake and maintain a flood alleviation strategy. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . for example. Where the community has religious or cultural attachments to the land. for example. The views. The first problematic area is that those where the community gains but the nation loses. the effect of the strategy adopted may simply be to increase the flood problems downstream. large numbers of people have to be resettled to make way for the project.
but not necessarily. the critical feature of dam is not its physical size but the distribution of its benefits and costs.World Commission on Dams Thematic Study of Flooding Figure 23 Distributional impacts of a project 105 B2 B1 area where project is to be built C1 C2 National net benefits = (B1 + B2) . and recommendations contained in the working paper are not to be taken to represent the views of the Commission . with its heavy reliance on the courts. the maintenance of social order is given as a primary objective. This assumed that there is an interest which is superordinate to individual or group interests and the problem is simply to find out the best means of reaching that goal. The views.g. This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. In the USA. large dams are typically. An explicit assessment needs to be made of the distributional impact of a project and an appropriate strategy developed in response. such as those being promoted by some of development orientated NGOs (e. to be found in upper right quadrant whilst small scale dams. Choices often therefore involve conflicts between different parties as well as objectives. In much of Europe. a more adversial approach appears to be typical. the use of the term ‘stakeholder’ hints that a more conflictual approach is adopted.C1 Table 20 Possible distributional outcomes Local net benefits National net benefits positive positive win-win situation negative transfer sufficient of national gain to local area to make local net benefits positive negative find an alternative approach? In relation to this form of analysis. Different societies define differently the overall objective that is to be achieved in societal decision making.(C1 + C2) Local net benefits = B1 . the basis for decision making is often cited as being ‘communal solidarity’ and China. the approach in the Netherlands has been described as highly consensual whilst in the USA. Different societies also bring different approaches to the search for achieving the societal objective. However. Development Alternatives 1999) are to be found in the upper left quadrant. One of the reason that choices have to be made is because no one option is superior in terms of its performance against all of the objectives that one or more parties bring to the choice. conclusions. The traditional British model was to determine the public or national interest.
option A can be eliminated. The policy in the USA for wetlands of ‘no net loss’ (Heimlich 1991). the best strategy for ensuring that all important objectives. if action is not taken one environmental resource will be lost or damaged but if action is taken. exemplifies this approach. In the case of environmental resources deemed to be ‘constant natural assets’. for example. Conversely. Over time.World Commission on Dams Thematic Study of Flooding 106 What is clear is both that there is more to life than economic efficiency and that there may be importance differences in the importance attached to the different objectives by different individuals or organisations. An alternative project appraisal technique. a progressively wider range of goals have been argued to be part of decision making. It quite explicitly excludes consideration of equity issues. Various graphical and statistical techniques can be used to explore the implications of making different choices. not least amongst economists. Figure 25 illustrates a hypothetical case where several options are being compared against a number of objectives. however. It is also based upon some strong ideological assumptions which may limit its relevance to. European Habitats Directive will allow damage to sites designated under it only when there are both no alternatives and overwhelming social and economic reasons. A significant weakness of benefit-cost analysis therefore is that it is limited to consideration of a single objective. such decisions should never be made by simply looking at the weighted mean scores of the different options. One such approach is to adopt the principles of ‘critical natural capital’ and ‘constant natural assets’ (Countryside Commission et al 1993). Here it is possible to narrow down the options to two: options B. Weighting the two options scores against the importance weights given to each objective implies that option C should be chosen. C and D all perform better against all the objectives than does option A. The extent to which can handle environmental issues is also somewhat problematic in that it based on a definition of value in terms of individual wants. are taken into account is to consult widely at the beginning of the project and continue to do so throughout the project cycle. So. no loss or damage can be accepted except in the most extreme circumstances. Thus. Multi-Criteria Analysis (Nijkamp et al 1990). typically involving mathematical operations that are not justified by the level of measurement achieved (Bisset 1978). However. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . The views. attention has been drawn to an increasing range of consequences that ought to be considered when comparing alternatives. Option C also performs better than option B against all the objectives and so the choice is between options C and D. preferably in advance. option C is at least as good as option D against all except the most important objective. equivalently. sets out to allow comparison of the options across as wide a range of objectives as interested parties bring to the choice. this option does not perform very well against the most important objective (number 5). Consequently. The extent to which it can encompass values given for moral reasons. Islamic societies (Deen 1990. Figure 24 Multi-criteria analysis: the analysis of alternative options This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. and therefore consequences. It is a good way of exploring and discussing preferences but a poor way of reaching any conclusions because too many important parameters are left implicit. conclusions. it generally suffers from a lack of rigour. These weaknesses do not reduce its utility as a means of exploring the trade-offs that a choice involves. coupled to wetland banking (Heimlich et al 1999). for example. is a matter of some debate. or through a concern for society. economic efficiency. another environmental resource will be lost or damaged. an individual site can be lost provided that equivalent is created. One way of then including environmental issues is as a set of constraints. Again. Nomani and Rahnema 1994). A choice must therefore be made as to which pattern is preferable. when. neither approach offers a way of dealing with conflicts between environmental resources. Where an environmental resource is considered to be part of ‘critical natural capital’. or.
which may range from developing a flood management plan for an entire catchment down to a project to manage floods within one small area. Therefore a mixture of techniques will often be appropriate. and necessarily the mean rank is then towards the average rank. then the different criteria yield different conclusions and a more considered assessment is necessary to determine what priority should be accorded to those basins. and critical strategic economic or transport links. the Government of Hungary recently sought to prioritise 151 flood basins for maintenance and rehabilitation (Halcrow et al 1999). The mean rank of each basin and the standard deviation of its rank were then plotted against each other. the different project appraisal techniques have different strengths and weaknesses.5 2 1. Across catchments it may also be necessary to determine what areas should be given priority for more detailed study. This is a working paper prepared for the World Commission on Dams as part of its information gathering activities.World Commission on Dams Thematic Study of Flooding OVERALL OPTIONS SCORES 5 4. environmental or heritage areas at risk of floooding. Finally. then all of the criteria agreed that the flood basin should be given a low priority. The individual basins’ ratings or other scores on each of the criteria were converted to rank order scores.5 0 1 2 3 O BJECTIVE 4 5 A B C D 107 Thus. those where they are definitely not. The appraisal led approach means that the appraisal process must be matched to the stage in the development of the design. with simple assessments being made at the pre-feasibility stage through to detailed assessments to confirm the selection of the option finally adopted.5 3 2. an assessment was made in the Netherlands (Peerbolte 1994) of the best programme for upgrading coastal dikes in response to sea level rise. For example. and those which need further consideration before a decision can be made. or whether the project should be abandoned now. One of the questions at the pre-feasibility stage is whether it is worth collecting the detailed data necessary before a detailed design can be developed. conclusions. no one technique being superior against all of the criteria that might be applied (Penning-Rowsell et al 1992). a number of different criteria were to be considered important including the risk to life from dike breaching. a triage model can be adopted which identifies those areas where detailed studies are definitely justified. Thus. a simple points system was developed for use by the Government of Ireland in prioritising 76 river catchments for more detailed study. For instance. if the mean rank was high and the rank standard deviation was low. It also requires matching the appraisal to the problem.5 4 3.5 1 0. the economic losses from flooding. In such studies. The triage approach adopted is illustrated in Figure 26. If the mean rank and rank standard deviation of each basin were both low then all the criteria agreed that that basin should be given a high priority. The views. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . In doing so. Conversely. if the rank standard deviation is high.
50 0.00 1. Rather. Figure 26 ratio Confidence in the economic viability of a project as a function of the benefit-cost This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. then this is usually a sign that a fundamental error has been made in the analysis. uncertainty as to what option to adopt. and recommendations contained in the working paper are not to be taken to represent the views of the Commission .00 0 2 4 m ean rank 6 8 108 definite priority basins definite low priority basins What this example also illustrates is that assessment can be defined as a process of converting data into information where the formal definition of information is that which removes decision uncertainty (Cherry 1966).50 std dev of rank 2. the benefit-cost ratio is a measure of our uncertainty as to whether the particular ‘do something’ is preferable to the baseline option.00 0.3 then we can be very confident that the baseline option is preferable to the ‘do something’ option and if it is 5 then we can be very confident that the ‘do something’ option is preferable to the baseline option (Figure 27).World Commission on Dams Thematic Study of Flooding Figure 25 Project prioritisation by multi-criteria analysis mean rank and agreement between criteria 4. if the benefit-cost ratio is 0. Finally.00 2.50 4.00 3. When considered in these terms.50 std deviation of rank 3. Instead of a benefit-cost ratio of one being a hurdle which if exceeded by a do something option demonstrates that that option should be adopted. if the benefit-cost ratio is greater than about 12. The views. conclusions.50 1. it is the point of maximum uncertainty as to whether the do something option is preferable to the baseline option.
Whilst the nature of the uncertainties affecting each parameter may be systemic.3 1 5 15 A second form of uncertainty to be managed. conclusions. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . and the formal definition of information (Cherry 1966) is that which destroys uncertainty. is between the different available ‘do something’ options. The views. Here. an approach which can be useful is to identify the differences in the parameters to which the individual ‘do something’ options are most sensitive (Figure 28). such a display may aid in determining which parameters are of greatest concern and hence upon which it is least satisfactory to rely upon in choosing the option.World Commission on Dams Thematic Study of Flooding confidence that cost ratio is greater 1 109 forget marginal do benefitratio 0. This is a working paper prepared for the World Commission on Dams as part of its information gathering activities.
The views. conclusions. and recommendations contained in the working paper are not to be taken to represent the views of the Commission .World Commission on Dams Thematic Study of Flooding Figure 27 parameters 110 Differentiation sensitivity of alternative ‘do something’ options to key Options A return period of threshold • event capital costs • maintenance costs effective scheme life rate of increase in runoff B C D • • E • • • • • • This is a working paper prepared for the World Commission on Dams as part of its information gathering activities.
project preparation studies. overall it was too ambitious. overestimating public sector capabilities and overemphasising large-scale surface water interventions. NWP Phase II was completed. Phase I of the National Water Management Plan (NWP) was completed. The compromise five-year plan that emerged from this debate was called the Flood Action Plan (FAP). and over half still depend on agriculture for livelihood. international commercial interests that favoured structural interventions. In 1991. and Meghna Rivers bring about 1x1012 m3 of water plus 500 Mt to 1500 Mt of sediment into Bangladesh from the upstream catchment area (area 1. FAP was strongly opposed by local and international NGOs. During the monsoon months (Jun-Sep). did not accept this plan. intensifying flood impacts and placing severe constraints on flood control options. and that planners could work in isolation from the people for whom the interventions were intended. conclusions. Rainfall within Bangladesh accounts for a further 0. Increasing population density and agriculture dependence compels people to inhabit floodvulnerable areas. Normal annual flooding provides numerous benefits . Over 80% of the population live in rural areas. a debate on how to address the flooding problem began to develop through various preliminary studies. After the 1988 floods. Though this Plan did led the way to protecting most of the coastal zone from tidally-induced flooding.common access to the large natural floodplain fishery. In 1964. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . flash floods affect smaller areas of the country that are located on hill streams and in piedmont areas. Floods are then slow to recede due to the low average gradient (about 5 cm km-1). To cope with these challenges. aggravate the situation. In 1986. and also. in part. It was overtaken by events when severe flooding in 1988 led to the formulation of the Flood Action Plan. This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. over the past several decades water resources planning has evolved in three phases: national water planning. and pilot projects. among these. causing catastrophic losses. and flushing of stagnant water in low-lying areas. The opposition to FAP challenged a number of basic assumptions related to structural flood control. These proposed interventions ranging from an almost purely structural “once-for-all” massive engineering solution.74 million km2).12x1012 m3. and drainage system conveyance limitations caused by sediment deposition. FAP consisted of regional planning studies. that the major rivers could be embanked sustainably despite large sediment loads and alluvial soils. But when the major river flood peaks coincide. unusually high floods can occur. the Ganges. the Flood Action Plan (FAP). if also present. even though population growth has been significantly reduced in recent years. that structural measures were affordable. planners emphasised ground water development for irrigation. The debate was subsequently short-circuited by a set of eleven principles prepared to guide future studies. organised around a coalition of environmental NGOs that initially raised awareness through public meetings outside Bangladesh. lacking implementation details. later the key to rapid irrigation expansion. This time around. Population has increased from about 70 million in the early 1970s to about 130 million in 2000. and post-FAP. This was primarily a food grain self-sufficiency sector strategy. directives from senior levels of Government to proceed despite the unresolved issues. It largely overlooked the country’s ground water resource. to a mainly non-structural “living with the floods” approach. including a detailed investment program. a national (at that time provincial) water planning approach was initiated with the 20-year Water and Power Master Plan. mindful of the weak performance of existing flood control drainage infrastructure. Adverse tidal conditions and heavy local rainfall. deposition of fertile loam on agricultural fields.World Commission on Dams Thematic Study of Flooding Bangladesh – The Evolution of Planning For Flood Control 111 Bangladesh (area 144.000 km2) experiences flooding every year on up to two-thirds of its territory. that flood control was desirable. The views. Brahmaputra. In addition to and distinct from the major river floods. concerned about possible overestimation of ground water. when 80% of annual rainfall occurs. with 172 million forecast for 2025. The Government.
rather. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . greater emphasis on participation has led to the recognition that as people’s lives are not compartmentalized by sectors.” the need for greatly enhanced people's participation. the FAP did not recommend large-scale works.World Commission on Dams Thematic Study of Flooding 112 The FAP process gradually produced a consensus on several issues. The views. Furthermore. and that cooperation among riparian countries is essential. It is now widely accepted that planning must be participatory and that consultation at all levels is essential to correctly identify development needs and interventions. flood mitigation as an integral part of flood management. so too must planning be multi-objective and multi-sectoral. among them support for a softer “controlled flooding” concept in place of the more hard-edged idea of “flood control. It is clear now that FAP resulted in a new planning approach. Other outcomes of FAP included much greater emphasis in the planning process on environmental and institutional aspects. In the end. conclusions. and an emphasis on improving drainage through dredging planned at the river system level. Herb Wiebe This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. acceptance that flood control should be addressed in a regional context. it initiated guidelines on people’s participation and environmental assessment.
corruption. There are. Looking backward is only useful to the degree that we learn lessons for the future. weaknesses in research. and an optimism about what the adopted solution could achieve. • has been arrived at through public involvement. • Technologically driven – the latest innovation was always taken to be the best and alternatives were frequently ignored. the past is largely a history of failure in water management and much of the current investment in those countries is devoted to seeking to recover from the damage done to rivers and catchments over the last one hundred years. conclusions. Barriers to sustainable flood management We can either look forward or look backward. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . the pressures to export solutions from one country to another. including its place within the catchment. Past approaches were characterised by: • Heroic solutions – a belief that the problem could be fixed. • is based upon adaptive management (Walters 1997). In the developed world. the difficulties of maintaining both physical and equally institutional options. • Simple objectives .World Commission on Dams Thematic Study of Flooding 113 7. an assumption of greater understanding as to the nature of the problem than existed. however. A sustainable flood management strategy is one that: • is appropriate to local conditions. • Static analysis – a static view of the problem and of rivers was adopted. experts and those with money. • Representing the interests of those with power – decisions were taken by narrow elites which included politicians. the desire to import the most modern solution.the objectives considered were narrow. The views. The most useful of those lessons of the past are often about attitudes rather than about the technologies adopted. and the problems of legitimising the involvement of different voices. This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. These are weaknesses that we can seek to avoid but the limitations of current knowledge will always be a problem. a number of major obstacles to be overcome before sustainable flood management can become a reality. professional roles. • A belief that resources were effectively infinite – as opposed to be scarce and requiring to be conserved for the most important purposes. the difficulties of enabling effective public involvement early enough in the decision process for this to be effective. These include: • • • • • • • • institutional limitations. and. equally. • and one that can be maintained.
Similarly. • constructor . their frequency and expected cost should be as much a part of a project design as the bill of quantities. the mid-river boundary is one of the most obvious contradictions to holistic catchment management and there must be some way of ensuring co-ordination across such a boundary. However. and whether it will be possible to generate the funds necessary to provide the level of maintenance the project requires. It will rarely be possible to redraw all administrative boundaries to reflect catchment boundaries. In consequence. Designers may therefore pay little attention to whether the system can be maintained. if there is to be local involvement and community funding.1 Institutional barriers In order that they be accountable. This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. The views. different institutions have been allocated responsibility for above water/ground works and those below water/ground. for example. Equally. conclusions. The administrative and community boundaries frequently developed down the centre of rivers.World Commission on Dams Thematic Study of Flooding 114 7. splits of responsibility between either or both construction and maintenance of bank protection and the dike have been introduced. or where there is considered to be a lack of adequate expertise at local level to successfully execute capital works. to natural hard points at the ends of a dike. one which is formally controlled through the contract document. • bank/dike. all institutions must have rules since we require that institutions perform predictably and consistently their designated role. Private contracting of works also introduces another form of split. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . it is necessary to determine where to draw both geographical and functional boundaries. Other types of boundary are introduced by funding practices. Thus boundaries are inevitable but at the same time they are the line along which major problems occur. there is difference between the institution made responsible for constructing the project and the organisation who will be responsible for operating and maintaining it. and • foundations/dike. Such a split is particularly likely when a higher governmental level is providing the capital funding. If the area and functional responsibilities of an institution are enlarged then the result is simply to convert the boundary problem into an internal problem. In some countries. For the same reason. but not Operations and Maintenance costs (O & M). how it should be maintained. Therefore. the problem is usually how to obtain co-operation across functional and geographical boundaries. In other cases. Figure 29 illustrates the different ways in which institutional boundaries have been drawn for managing dike systems: • across river. A maintenance schedule spelling out the actions. all institutions must have both geographical and functional boundaries: these define what they can do and within what area. the institution having to determine how to split itself up into areal and/or functional responsibilities. different institutions will have different appropriate geographical boundaries and these are commonly the result of historical. administrative and community boundaries are unlikely to correspond.operator. • down river. quite often in water management. then the same boundaries will generally have to be followed. cultural. ethnic and other factors.
and neither are stationary. such as the Dutch ‘Stork’ plan (de Bruin et al 1987). it is essential to devise ways of bridging the necessary boundaries. However. Communication across boundaries is generally problematic. However. the problem is even more complicated because the international boundary may lie mid-river or along the Thalweg. Because there are no ideal boundaries between institutions and their responsibilities.World Commission on Dams Thematic Study of Flooding 115 For rivers. conclusions. The views. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . One of the key rule structures governing institutions is how they may raise and spend money. the failure of flood warning in the 1997 Red River Flood along the North Dakota/Minnesota border is an example of such breakdowns across institutional boundaries (Pielke 1999). budgets are frequently single-functional and funding a multi-functional project will frequently require putting together a coalition of institutions whilst at the same time getting them to agree how the costs are to be shared. The latter approach is generally more likely to be effective since lower order institutions frequently spend a great deal of their time subverting the decisions of a super-ordinate institution and the superordinate institution tends to become a scientific bureaucracy inventing paper plans unrelated to conditions on the ground. this rule is typically set out in the legislation establishing an institution. such as the Mekong. The remedy is to switch away from rules to greater oversight by the legislatures. One consequence of holistic catchment management is that multi-functional projects will be increasingly common. cooperation and communication between institutions are necessary if integrated water and land plans. Nor was this failure atypical and flood warning systems are particularly prone to failures in communication across institutional boundaries. such as a catchment management agency. This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. Multi-functional projects require multi-functional funding but increasing the flexibility an institution has in spending money increases the risk that it will mis-use its budget. that form an international boundary. are to achieved. or • establish a new co-operative institution largely or wholly ‘owned’ by the existing institutions. Two possible approaches are: • establish a new super-ordinate institution.
This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. public involvement towards geographically small. Therefore. If an institution focuses entirely on avoiding making mistakes. A reason for creating institutions and establishing rules is to increase accountability through routinising decisions: if the rules are accepted and correctly applied so it follows that the decision is appropriate. the more easy it is for the institution to reply to any criticism that we followed the rules. Both institutions themselves and society also need to learn that it is acceptable for an institution to say either that they made a mistake or that they don’t know the answer. some mistakes will inevitably be found in hindsight.World Commission on Dams Thematic Study of Flooding 116 Figure 28 The different ways in which institutional boundaries have been drawn with respect to dikes administrative boundaries are typically down the centre of the river end of dike section bank protection versus dike protection foundations and underwater works versus dike structure constructorowner An apparent contradictions at the heart of the criteria for sustainable development set out in both the Dublin Declaration (ACC/ISGWR 1992) and the Rio Principles is that between holistic. or integrated catchment management. Integration tends to pull towards multi-functional organisations which cover the entire catchment. and the tighter the rules. conclusions. Decision making is not risk free. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . community based involvement. The staff of the institution know where they stand: doing the same thing that has been done for the last ten years then becomes the safe solution in terms of accountability. Requiring accountability thus has the perverse result of stifling innovation and institutional adaptation: rules are necessary for accountability. Where the rules are open to interpretation. knowledge and concerns. This should be resolved by catchment agencies which are downwardly accountable and which act to promote agreement between these smaller and more specialised groups as to the overall strategy to be adopted. it will never innovate. accountability by open review rather than over-restrictive rules is preferable and institutions need to incorporate advisory groups who will bring in diverse perspectives. the easier it is for the institution to be criticised after the fact and the usual consequence is then a demand for the rules to be tightened to reduce the scope for interpreting the rules. The drawback with routines and rules is then that the same solutions can become applied to all decisions and the institutions do not adapt to changing conditions. and involving the public in all levels of decision making. The views.
to pass a new law or introduce a new tax. say. they need to learn from the mistakes made by the countries of Europe and North America rather than to replicate them. when giving development aid. often quite defined ones. This approach clashes directly with involving the public because.World Commission on Dams Thematic Study of Flooding 117 7. To achieve this. countries like Bangladesh. for public involvement to be meaningful. Before that expertise can be drawn in. engineers become engineers to build things. The views. However. they also believe that they knew best. governments have tended not only to export the current flood management practices of that country. Unlike many of the disciplines who will be drawn into such an approach. because of their expertise. Engineering training needs therefore to be orientated towards training engineers to listen to. However.2 Professional roles People train in the professions because they want to perform what they believe to be the role of that profession. it involves finding out what the public want. This is rather different from the usual call for multi. Thus. particularly water engineers. and what lessons can be learnt.3 Exporting/importing failure Logically. Thus. conclusions. as upon identifying examples of good practice. 7. the need for that expertise must first be recognised and engendering such a pluralistic approach is part of the necessary training of engineers. if an ecologist were to propose a massive dam building programme on the grounds of employment generation but less so if the same programme were to be proposed for this reason by an engineer or an economist. the organisation has already prescribed the type of solution that is expected. but not necessarily to build some thing. when a problem is handed to an engineer there is at least an implicit expectation that the appropriate response is to construct some thing rather than. Engineers. for instance. India and Japan should become the global centres of expertise in flood management simply because they have bigger problems than anyone else. and also to put the interests of national companies ahead of the recipients. China. This is a development that must be promoted with engineering training being increasingly orientated to a wider problem solving approach into which they draw the expertise of other specialists. as to the role of particular professions. The aim should be skip an entire generation of mistakes. This means that study tours to the developed countries should be concerned at least as much as learning what were the mistakes than have been made there. In particular. have also developed an ethos of public service. when an organisation appoints a particular specialist to deal with a problem. This is expressed in the form of approach of determining what the public needed and then building it. society has expectations. Equally. We would surprised. These expectations extend on both sides to what the professions do. what should be done to promote the public good. Whilst in this way they put their expertise towards the public good. irrespective of their relevance to the recipient. although that will result from this approach. engineers are probably amongst the most multi-skilled professions and they are increasing proposing strategies which do not involve building things.4 Financing This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. The weaknesses of the sciences tends to be that they want to study some thing and are frequently more parochial in conceptual outlook than engineers. learn from and engage in a dialogue with the publics so as to assist the publics in achieving their goals. 7. This may be quite different from what the engineer thinks they need.and inter-disciplinary working. engineers have the advantage of a problem-orientated approach: we will still want them to do some thing. and recommendations contained in the working paper are not to be taken to represent the views of the Commission .
Therefore in practice what has happened in the past. is essentially an equivalent of the French concession approach in water supply and sewerage (Rees 1998). Secondly. But it is less easy to charge those who cause an increased risk of flooding through emissions of greenhouse gases. there are equity concerns since sole reliance on the principle would result in the rich undertaking flood alleviation measures and the poor remaining exposed to flooding. President Lee Teng-hui said: “To take care of the people is the responsibility of the government. so we will adopt an aggressive attitude and try our best to respond to this situation” (President Lee Teng-hui of Taiwan. following the recent earthquake in Taiwan. conclusions. The ‘user pays’ principle is as much a moral claim as anything else and can conflict with concerns both with equity and economic efficiency. practical and efficiency concerns. In some urban areas. therefore. One obvious problem is that these works have to be funded through what is essentially a tax and taxation decisions are reserved as the responsibility of some accountable governmental body rather than to a profit maximising private company. is that capital costs are at least partly funded by the central government.World Commission on Dams Thematic Study of Flooding 118 The debate as to who should pay for flood alleviation is a somewhat confusing mixture of ideological. Fourthly. Firstly. Thirdly. This procedure. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . such charges are applied. This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. Fifthly. Finally. Thus. where occupancy of the floodplain has been supported by a planning decision that this is the best place for intensified development to take place. and largely continues. the Private Finance Initiative. The views. In the UK. it makes little sense to then encourage it to take place elsewhere through differential pricing. flooding is to a greater or lesser extent an externality of land use. In principle. The Environment Agency will pay an annual sum rather than the works being funded through a specific local tax on the beneficiaries. There are a number of reasons why reliance wholly upon the ‘user pays’ principle to fund flood alleviation can conflict with both the latter objectives. Guardian 22/9/99 p15). two projects are being constructed and maintained through private funding. the extent of this funding depending on the relative wealth of the area where the alleviation project is to be undertaken. flood alleviation is technically a public good so that economic efficiency is not served by pricing some possible users out of the area. any changes in land use which increase the flood risk elsewhere should attract a charge. There is little successful history of the use of private capital in promoting flood alleviation works. in many countries there is a doctrine of communal solidarity and a belief that a government has a duty of care to the citizens of that country. the theory of ‘second best’ (Lipsey and Lancaster 1956-57) warns that seeking to correct one form of market failure where there are many failures can make things worse. occupancy of the floodplain is only one of the possible externalities of development.
The views. However. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . the charges levied by the Waterschappen in the Netherlands and similar charges in China being examples. Systems whose O & M requirements are capital intensive. there are a number of common problems. Too often there is a divorce between the designers of a project and those who then have to maintain it. for example. and • Inappropriate requirements in the way of required skills. The result is that the This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. better decisions will only emerge from better research. the Ministry of Finance will have a rooted objection to such an hypothecation of what is essentially tax revenue.6 Research In the longer run. is that the adoption in the past of structural engineering solutions and dams in particular has been a reflection of doubts. equipment and materials. the critical scarcities are typically in terms of equipment and skills. This is equally true of the failures of institutional structures as it is of physical structures. Emergency plans. again largely at the cost of central government. However. Conversely. More water related projects appear to fail from a lack of or insufficient maintenance than for any other reason. Firstly. in the short run. and it can be no more. 7. One way of securing funding is designated local taxes on local communities to pay the costs of O & M. skill intensive and rely on imported parts are inappropriate in countries where capital and skilled labour are scarce. The weakness of this approach is that local government may not then adequately fund O & M activities so that a renovation scheme must be undertaken prematurely. The high discount rates typically used in developing countries tend to result in the selection of low first cost and high O & M costs options over those which require lower O & M costs at the expense of higher capital costs. as in some parts of Bangladesh. Reviews of experience in rural water supply where there has been much more experience are not very encouraging (Black 1998). local residents may breach dikes because they blame the dikes for causing local flooding and regard the dikes as being imposed upon them by an alien government. The potential advantage of this split is that it provides local ownership of the project. the response is that more research should be funded first. Institution maintenance is an area for which only limited guidelines have yet emerged (Ostrom 1990). Thus. local residents also have a duty to turn out for emergency flood fighting and/or to provide unpaid labour for maintenance activities. It can further be argued the cause of insufficient maintenance of physical structures is a consequence of a failed institutional system. as to whether to institutional approaches could be maintained. universities continue to be centred around disciplines whereas problems obey no such rules. the designers failing to consider the O & M requirements of the project and whether they can be sustained.5 Maintenance There are at least three different reasons why projects fail as a result of a lack of maintenance: • Inadequate consideration during the project design of the ability of the managing institution to raise sufficient income to cover O & M costs. Operational and Maintenance costs are generally the responsibility of the local government in the area. • Inappropriate institutional design. conclusions. In both those countries. overseas debts restrict imports. and rates of pay are low in public organisations so that they have a problem retaining trained staff. frequently when the decision-maker asks a scientist what should be done. justified or not. scientists want to study things whereas decisionmakers have to do some thing. need to be rehearsed and the networks of organisations which will execute them need to be maintained.World Commission on Dams Thematic Study of Flooding 119 7. in most countries. Secondly. A hypothesis. where this does not occur.
World Bank 1996). The views. in effect. 7. the Environment Agency in England and Wales still. This militates both against multi. institutions should be accountable downwards rather than upwards. against applied research. If the institutions taking the decisions remain as scientific bureaucracies. At different times. the answer seems to lie in institutional structures and devolving decision power rather than simply improving communications. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . A danger that the process of public involvement will be captured by specific social groups. Where the This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. Computing developments may offer a way whereby the different publics can explore data. conclusions. talks in terms of ‘scheme promotion’. for example. A second problem is that public involvement can only be effective where the public have access to adequate information. particularly by the richer and more socially empowered members of society. Whilst it is not clear that we yet know how to achieve real and effective public involvement. women are to a greater or lesser extent excluded from such involvement. and also the dissemination of research findings to decision-makers. The two key areas remain institutional and information. so too are those who do not live in the area are present but might do so under particular conditions. and gain by so doing. In particular. Big engineering works arguably offer most scope for corruption by national or provincial politicians. Both money and prestige have proved quite effective ways of convincing academics of the desirability of a greater orientation towards the problems of society and the active dissemination of research results to those who need them. Thus. The ‘Landcare’ movement in Australia perhaps may be the most successful effort of long term public involvement to date. in most countries.World Commission on Dams Thematic Study of Flooding 120 decision-maker has to diagnose the problem to determine what is the appropriate discipline to call upon. its own framework of discourse. Thirdly.8 Public involvement Public involvement is still in its infancy in most parts of the world. A ‘Freedom of Information’ Act is not sufficient for this to occur since this will liberate a large quantity of data and data only becomes information when it is structured. a better model for a catchment management agency being one where the Board is made up of representatives of those who live or use the river rather than one which is responsible to central government and appointed by them. Instead. since any way of structuring data imposes its own view of the world. A number of handbooks have been prepared on how to increase public involvement (Creighton et al nd. this discipline orientation can result in an inward orientation wherein only the advancement of the discipline through the acclaim of peers is considered to be the appropriate role of an academic. of selling its preferred solution to other bodies.or inter-disciplinary working. concerns have been expressed that both Environmental Assessment and public participation can be used by parts of the local population to maintain the status quo and to exclude others. The requirement is to allow the publics to be able to structure what are often large quantities of data in their own way. Institutional approaches then have the disadvantage that the scope for corruption is at the local level rather than the national level. and impose their own structure upon it. academics may feel that it is up to the end-user to find out what is the latest research that bears on the decision. Future generations are necessarily excluded from such involvement. then public involvement is likely to be limited to consultation at best.7 Corruption Corruption continues to be a major problem and one that often influences the choice of strategy adopted since that strategy tends to dictate both how much money can be creamed off and by whom. 7. and in some countries this has provided a means of campaign financing.
9 Legitimising voices Involvement in decision making is also power. 7. this may be morally acceptable. In consequence.World Commission on Dams Thematic Study of Flooding 121 public involved are an indigenous group. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. they must be able to demonstrate that they are concerned with the sustainable livelihood of those who will be affected by the decision to be made. Secondly. where it is the relatively rich who use the system to exclude the poor. A further perceived problem with increased public involvement is that it is feared that the whole decision process will be slowed down if not actually grind a halt in an interminable process of consultation. all those seeking entry to the decision process have to answer for their legitimacy: by what right are they entitled for their voice to be heard in a decision? At a minimum. it will be resisted. conclusions. The views. inclusion of an additional voice in the decision process will almost certainly be seen by one or more of the existing decision makers as resulting in a diminution of their power. it is less so.
Similarly. and • Adopting a catchment management approach. What then emerges as the best option or the best combination of options for flood management then will depend upon local conditions. conclusions. Flood management in North America and Australasia should consequently be much easier than in the rest of the world. we appear to be much less capable of developing sustainable institutions and thus of implementing institutional approaches to flood hazard management than of implementing engineering works. More generally. Flood warning systems are most likely to fail when conditions are unusual. in terms of the available flood management options. They are not ways of replacing deliberative decision making through public involvement. high availability of arable land and low intensity of economic activity usually mean that the trade-offs are less intense and there are more options available than in the rest of the world. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . But they are useful only to the extent to which they promote understanding and enable that understanding to be shared. That 30% of hand pumps in India are estimated to be out of commission is largely down to the failure to devise or implement appropriate institutional forms. The overall objective must be sustainable development. The views. The conditions under which land use and building controls on floodplain use can be effective are still to be established but they are unlikely to effective in those areas where they are most needed: in areas of rapid urbanisation. decisions are difficult precisely because there is no course of action that is better in all senses than all other options. however. or rather the storage that can be provided. Because decisions are complex. Brundtland went on to assert that sustainable development requires the reductions in the present inequalities of wealth between peoples. In the long term. However. flood warnings work best where floods are frequent and where arguably flood warnings are least needed. The widely quoted Brundtland definition of sustainable development recognises our duties to future generations. the developed countries can expect to make more progress towards this goal than those countries where population densities are high. Equally. Conclusions and Recommendations Dams. in those extreme floods where they are most needed. there are painful trade-offs involved in choosing between flood management options. particularly where much of that development is via informal settlements. In the short term. the low population densities. It is the nature of decisions that making a decision involves trade-offs between important objectives or values. The Dublin meeting and the Rio conference set out the guiding principles for making decisions so as to achieve sustainable development as being: • Maximising public involvement at all levels of decision making. When such approaches are implemented then outsiders must accept the decisions that result from the approach even when we disagree with the decision itself. the use of decision aids and tools such as benefit-cost analysis and multi-criteria analysis have an important role to play in clarifying the nature of the choice that must be made and understanding the critical trade-offs involved. however. they are simply means of informing that process. for instance. Both are necessary conditions for sustainable development. In the countries of nineteenth century settlement.World Commission on Dams Thematic Study of Flooding 122 8. only generalisable principles. Much flood management may then have to be This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. can reduce flood losses and have an identifiable role to play in some circumstances. There are no universal solutions. it is no use being against one option unless there is a better feasible option. restoration of rivers to a more natural form and of their associated wetlands is a widely held objective. Generally. there is a shortage of arable land and levels of poverty are unacceptable.
The specific recommendations were summarised in the Executive Summary. Decisions about the most appropriate flood management strategy to adopt must consequently both take account of the effects of climate change and the contribution of the flood management strategy to slowing climate change and then stabilising the climate. We are at present. flow and temperature conditions in those streams will render them incapable of supporting trout. no country either. seeking to restore trout streams whilst there is a significant probability that in fifty years time. in some cases. Climate change will make appropriate water management more difficult in all countries and climate change is probably the largest threat to biodiversity. If we do not cut greenhouse gas emissions then in fifty years there will be no biodiversity left to be impacted by flood hazard management and. conclusions. This is a working paper prepared for the World Commission on Dams as part of its information gathering activities.World Commission on Dams Thematic Study of Flooding 123 seen as a holding strategy until the country concerned achieves a level of development that is comparable to those in countries that are currently considered to be developed. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . The views. for example.
References and bibliography ABI/CGU Insurance. Army Corps of Engineers.A. T. 1992. Simpkins. VA. J. conclusions. ecosystems functions and environmental restoration. The Dublin Statement and the Report of the Conference.. and Madiec. Taylor. Water Resources Bulletin 25(4).Y: Simon and Schuster. Osgood. (eds. M. “Montpellier met a l’eau sa future universite”. A. and George. 1999. E... C. and Hollis. Paris: LATTS-ENPC Barry. report to Anglian Region. J. Birley. H. B. Fort Belvoir. S. “ An evacuation model for major accidents”.. NewYork. 1994. Dams. 1999. NPPG 7. Bellamy.. M. Balfour Maunsell. 1992. P. M. A. Behavioral and Physiological Effects of Chronic Stress at Buffalo Creek”. D.and Insurance. and Schaefer. October. P. Disasters and Emergencies: the Need for Planning. Motz. Flood Appraisal Groups. J. A. Jenkins. 881-890. 1994. 50 Millions de Consommateurs No 265. M. Cape Town: World Commission on Dams (http://www. and Love. The views. . N. B. Baum. 1995.. Abt. R. N. J. P. D. D. J. B. D. “Determination of Urban Flood Damages”. Dams and Floods. Natural Disasters London: UCL Press Limited. and Gershuny. London: HMSO ARSIA/GEOPLAN/CEHIDRO/WASSER WIRTSCHAFT.. “Environmental Effects of Hydro-Electric Power Generation in Africa and the Potential for Artificial Floods”. 1993. River Nene Structures .org) This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. Oxford: Oxford University Press Anon. Rising Tide.CIWEM 10. 1988. Antle. and Westbrook. Institute for Water Resources Anderson. C. Barbier. 1 June 1992. 1985. paper given at the Workshop on Benefits of and Concerns about Dams. 1997. J. British Medical Journal.) 2000. 429-435 Acreman. London Bennet. the Experience in Bordeaux (France)” Paper presented at the CONFLO 92: Integrated Catchment Planning and Source Control.. Paris: ICOLD Berga. Gatchel..: U. (eds.I. M. 1979.D. Gland: IUCN Acreman.. EUROWATER: Vertical report on France. O. Wallingford: Institute of Hydrology Adams. J. A. 2nd Edition. report to DFID.. E.S. Dugan. R & D Report 12.S.a review of current problems of current problems and future prospects. Geneva: World Meteorological Organization Acreman.J. et. Journal of Consulting and Clinical Psychology 61(4) 565572. Wasting the Rain: Rivers. “Benefits of Dams in Flood Control”. and Cambon. December. Managed flood releases from reservoirs . F. G. Berga. Chateau. F.. (ed. “Bristol floods 1968: controlled survey of effects on health of local community disaster”. C.. Project R7344. W.J.) 2000. M. Anatalya Bergkamp. L. Berland. J-M. G. The Political Economy of the Household...World Commission on Dams Thematic Study of Flooding 124 9. L. (eds. Journal of the ASCE. 1993. W. L. W. Water Resources Division 111(3). 1989. Oxford UK.) 1994. People and Planning in Africa. Wittler. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . 22-26 Appelbaum. G. Allee. National Rivers Authority Barraque. E. S. 269-282 Arnell.) 1996. “Human stability in a high flood zone”. 1999.M.dams. ACC/ISGWR 1992.. and McNeely. 1980 Human Costs of Flooding and Implementability of Non-Structural Damage Reduction in the Tug Fork Valley of West U. PREMO 98 Prevention in the mountains for protection of the valleys Pisa: ARSIA Balades. L. al. “ Emotional . The Implications of Climate Change for the National Rivers Authority. A. Water Management and Wetlands in Sub-Saharan Africa. S. and Bechhofer... Perth: Association of British Insurers/CGU Insurance. 1983. August 21. M. Oxford University. 1999. London: Earthscan Alexander.J. 454-458. “A Means of Fighting Urban Storm Water Overflow: Compensating Techniques. paper presented at the Conference .. 1996. Van der Slice. R. G.J.R. and Harrison.benefit assessment.
Nepal. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . 1988. Bonn: Bundesministerium fur Ruamordung. Brookes A 1990. Brookes.. S. Applied Geography 6. Oxford: Blackwell This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. “The effect of forestry on the quantity and quality of runoff in upland Britain” in de L. Blaikie. H. Chichester: John Wiley. Bristol: the Environment Agency Calow. Bruins. I. Bureau of Reclamation 1988. J. and Petts. and Disasters. Anatalya Binder. decision-making and environmental impact assessment in the United Kingdom”.Solbe.. Schulter. G. U. “ A Review of Catchment Experiments to Determine the Effect Of Vegetation Changes on Water Yield and Evapotranspiration” Journal of Hydrology 55. Chichester: Ellis Horwood.) 1994. Planen and Bauen von Gebauden in hochwassergefahrdeten Gebieten: Hochwasserschutzfibel. B. The Nature of Structural Design and Safety. and Lindstrom. Bonn: Bundesministerium fur Bildung. Blockley. flood mitigation and environmental aspects”. Research Conception: Ecological Research in the Elbe Catchment Area. Black. Highland-lowland interactions in the Ganges Brahmaputra River Basin: a review of published literature. River Channel Restoration: Guiding Principles for Sustainable Projects. 13-32 Bundesministerium fur Bildung.. Regulated Rivers 5.A. “Floods in Regulated Rivers and Physical Planning in Sweden”. Wissenschaft. 1980. Bauwesen and Stadtebau 1996. Schutz der Binnergewasser. Report by the Independent Review Team to the Board of the Environment Agency. S.. M. 1994.D. At Risk: Natural Hazards. M. Davis. (ed) Effects of land use on Freshwaters. 1996. R. Geographical Journal 156. and Hewlett. D. S. R. 1978-1998 Learning What Works: A 20 Year Retrospective View on the International Water and Sanitation Cooperation.D. H 1990. International Centre for Integrated Mountain Development.G. The Rivers Handbook: hydrological and ecological principles. People’s Vulnerability. 389-412. R. Brammer. and Bremmer. Forschung und Technology Bundesministerium fur Ruamordung. P.A. 1992.J. CO: Bureau of Reclamation. and Horner. and Nessler. 1989. “Floods in Bangladesh 2. 1986. Journal of Environmental Management 7. Washington DC: UNDP-World Bank Water and Sanitation Program. ACER Technical Memorandum No 11. D. and Newsom.F. Channelised Rivers: Perspectives for Environmental Management. M. 1999. and Wagner. Black. “Rainwater-harvesting agriculture for food production in arid zones: the challenge of the African famine”. Dundee: University of Dundee.. volume 2. I. P. 2000. D. G. A. Wissenschaft. R. 3-23. Jurging. 1990. W. Bunza. Hydrology of the Moist Tropical Forest and Effects of Conversion: a State of Knowledge Review. Vol. Evenari. D. 158-165 Brammer. 43-58 Black. P. Bonn: Economica Verlag Bisset. I. (ed) 1998. L. J. H. L. T. 1998. “ Flood Vulnerability and Flood Disasters in Bangladesh”.. A. Bruijnzeel. J. Cooperating with Nature: Confronting Natural Hazards with Land-Use Planning for Sustainable Communities. Forschung und Technology 1995. ICIMOD Occasional Paper No 11. 1994. 1999. C. G. E. 1978. “Restoration and enhancement of engineered river channels: some european experiences”. “Quantification. (ed) Floods London: Routledge. paper given at the Workshop on Benefits of and Concerns about Dams. Denver. and Evans. D. F. (eds. Flood damage in the UK: new insights for the insurance industry. Flood proofing rural residences. UNESCO IHP Humid Tropics Programme.I & II. Wisner. W. and Shields. Chichester: Ellis Horwood Bosch . Netherlands. Canon. conclusions. in Parker. J. 1986. Economic Development Administration Blackie. The views. Easter 1998 Floods. Chichester: John Wiley Bruijnzeel. Bauwesen and Stadtebau Burby.M. Engelhardt. Washington DC: Joseph Henry Press. 1975.R..N. Downstream hazard classification guidelines. 45-56 Brookes. Kathmandu.World Commission on Dams Thematic Study of Flooding 125 Bergstrom. A. report to the US Department of Commerce. London: Routledge. P. J. U. 1998. M. US Department of the Interior Bye.
.T. S. C.... “Vulnerability Analysis and Disasters”. 1993. Overmars. D. 1997. H. (ed) Floods London: Routledge. 1966.. van Nieuwnhuijze. Enfield: Middlesex University Chan. CIRIA 1999a. D. “Response to Dynamic Flood Hazard Factors in Peninsular Malaysia”. B. Enfield: Flood Hazard Research Centre Cherry. J. W. J. Strategic Land Use Planning in Europe and USA: Review of the Best Practice for the Environment Agency. H. L. Anatalya Darby.. 969-96. Benefit-cost analysis of Hydrometric Data . 1996. L. Environmentalist 7(2). for the Department of the Environment.A. C. D. and Ayres. C. B. Committee to Review the Glen Canyon Environmental Studies 1987. Conservation Issues in Strategic Plans. London: Belhaven This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. V. Priscoli. 2000. Geographical Journal. Y. “Environmental function as a unifying concept for ecology and economics”. Sustainable Urban Drainage Systems Design Manual for Scotland and Northern Ireland. G. M.. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . report to National Rivers Authority Welsh Region. S. Sustainable Urban Drainage Systems Design Manual for England and Wales. Boulder. The views.Best Practice. 2000. F.. “Role of Reservoirs in 1998 Heavy Flood Fighting on the Yangtze River”. D. “ The Record 1993 Missippi River Flood: a Defining Event for Flood Mitigation Policy in the United States”. 1983. J. W.W. Cambridge: Cambridge University Press De Bruin. De toekomst van het rivierengebied. Rotterdam: Balkema Chan. Changnon. On Human Communication. 1999. Technical Report W62. 162. 1993. A contextual analysis of flood hazard management in Peninsula Malaysia. and Parker. 1992 Principles of Water Law and Administration. Marlow: Foundation for Water Research Collins. Northampton: Countryside Commission. Ooievaar. E. “Evaluation of forest practices prescriptions from Washington’swatershed analysis program”. 1991. Swindon: Water Research Centre Chatterton. D. and Pess. C523. 1987. M. Dunning. law. 134-70. CIRIA 1999b.35 (5). 3: 313-325. House of Representatives. C. Wentlooge Levels Benefit-Cost Analysis. S. N. B.D. in Parker. Public involvement and dispute resolution: A Reader on the Second Decade of Experience at the Institute for Water Resources. T. J. US Army Corps of Engineers Czerwinski. 1995. Green. CNS Scientific and Engineering Services 1991. Changnon. MA: MIT Press Chow. Alexandria VA: Institute for Water Resources. London: CIRIA. J. 1990. International Association of Scientific Hydrology 42. “Hydrologic studies of floods in the United States”.R. Cambridge. I. and Vera. D. Impacts and Responses. London: CIRIA. W. Washington DC: National Academy Press Countryside Commission/English Heritage/English Nature.D. 1956. C521. and society” in Engel J R and Engel J B (eds. 1996. The changing Fenland. N. D. Caponera. “Islamic environmental ethics. Flood Insurance: Information on Financial Aspects of the National Flood Insurance Program.) Ethics of Environment and Development. River and Dam Management: A Review of the Bureau of Reclamation’s Glen Canyon Environmental Studies. Vol. R. unpublished PhD. and Siping. C. Co: Westview Press. and Green. 1999. (ed). Testimony before the Subcommittee on Housing and Community Opportunity. S. Arnhem: Stichting Gelderse Milieufederatie de Groot. 1987. London: CIRIA. C. Journal of American Water Resources Association.J. and Penning-Rowsell.river flow gauging. D. Hamhuis. H. in Parker. 1999. The Great Flood of 1993: Causes. (ed) Floods London: Routledge. Sijmons. Creighton. Chatteron. CIRIA 1999bc Sustainable Urban Drainage Systems . C522. Washington DC: United States General Accounting Office Daoxi. paper given at the Workshop on Benefits of and Concerns about Dams.World Commission on Dams Thematic Study of Flooding 126 Cannon. 105-109 Deen. A. conclusions.A.
Mount Macedon. and Sherman. T. in Balabanis P. 1995. and Pedroli.C. Washington. 1987. Washington.E. “Flood control and ecological rehabilitation of the Meuse River in the Netherlands”. Condition Assessment Manual: A Guide to the Visual Condition Assessment of Sea and River Defence. San Francisco: Freeman. Technical Annex14 for the Evacuation Module. “Social Factors that Constrain Human Responses to Flood Warnings. Center for Agriculture in the Environment (http://www. in Balabanis P. B. Guide 3.edu/hazards) Drijver. J. Boulder. A. D. in Parker. (ed) Floods London: Routledge Drescher. Emergency Action Planning Guidelines for Dams. unpublished MPhil thesis. University of Leiden Duel. Carpenter.. P. A. The Politics of GM Food: Risk.The River Nahe Catchment”. London: Earthscan Drabek. M. and Abueg. American Farmland Trust. Economic Analysis of Environmental Impacts. Evaluation of FDMM and Guidelines for the Justification of River Maintenance. 1999. Report on the Costs and Benefits of Natural Hazard Mitigation.) Ribamod: Proceedings of the final workshop. T. J. Psychophysiological indicators of PTSD following Hurricane Iniki: the Multi-sensory Interview. Science and Public Trust. 1985. and Widstrand. The Professional Emergency Manager. EUROFLOOD:Evacuation. Australia: Australian Emergency Management Institute Environment Agency nd. Cranfield University Dunne. Dam Safety: An Owner’s Guidance Manual. M. “Population and Water Resources: A Delicate Balance”. and Leopold. 1998. Canberra ACT: EMA. P.World Commission on Dams Thematic Study of Flooding 127 Delaney. T. Taming the Floods: environmental aspects of floodplain development in Africa. Flood-proofing Nonresidential Structures. Washington D. 1992. 1998. L. Boulder CO: Natural Hazards Center (http://www. Water in Environmental Planning. F. A. “Integrated Flood Management . C.org/cae) Demuth. T. Bristol: Environment Agency Eriksen. “Benefits to Downstream Flood Attenuation and Water Quality As a Result of Constructed Wetlands in Agricultural Landscapes”. Washington DC: Population Reference Bureau Federal Emergency Management Agency 1985. G. Population Bulletin. Drabek. D.E. and recommendations contained in the working paper are not to be taken to represent the views of the Commission .E.2000. The views. B. Leiden: Centre for Environmental Studies. 1978.. A.C. N.. H. G. Falkenmark. Creating flood disasters. Casale R and Samuels P (eds. Ivens. and Marchand. Bronstert A.) Ribamod: Proceedings of the final workshop. 1986. Federal Emergency Management Agency. D. EUR 18287.B. Federal Emergency Management Agency.L.farmlandinfo. Casale R and Samuels P (eds. Silsoe: Silsoe College. L. New York: Springer Verlag. Scura. Bronstert A. Wellington: National Water and Soil Conservation Authority ESRC Global Environmental Change Programme 1999. 1986. EUR 18287. 1997. C. Human System Responses to Disaster: An Inventory of Sociological Findings. A. conclusions. Emergency Management Practice. Homeowner’s Guide to Retrofitting. Managing the Floodplain.. EUROflood 1996. Glas. 1995. M. 1995. E. Flood Warning: an Australian Guide. Drabek. Enfield: Flood Hazard Research Centre. Australian Emergency Manuals Series. R. F. D. Emergency Management Australia 1999. M. K. 1999. A.colorado. 5: University of Sussex. R. Luxembourg: Office for Offical Publications of the European Communities Dunderdale. J. Colorado: Institute of Behavioural Science. N. 1986. Camp George West Golden CO: Colorado Department of Public Safety Dixon. J. Luxembourg: Office for Offical Publications of the European Communities Department of Health 1997.C. T. London: The Stationery Office Division of Disaster Emergency Services 1986. Health survey for England 1995. 1994. This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. Special Briefing No. Quick Response Report #77. FEMA 64 Federal Emergency Management Agency. Emergency Management Australia. Water and Soil Miscellaneous Publications 77.
1995.M. and Penning-Rowsell.. H. “The economics of floodplain use”. EPA. H. H. J. C.. D. and Morrow. Floods . Garrido. Guidelines for Risk and Uncertainty Analysis in Water Resources Planning. Schaad. The idea of a flood. Fordham.ca. I. 1. Washington. Harmancioglu N and Yevjevich V (eds. Zurich: Swiss Re (http://swissre. The views. 1998. Menzinger. Acts of God and acts of man: Recent trends in natural disasters and major industrial accidents. Edinburgh: Forestry Commission.com) Gaschen.. IWR Report 92-R-1. D. H. Assessing vulnerability to flooding. W. 1982. Washington DC: Island Press Graham. Golding.) Coping with Floods. 1986. Enfield: Middlesex University. Paris: SCOR Green. S. Enfield: Flood Hazard Research Centre Green. E. 1992. “Inondations: gestion des zones inondables et perspectives pour l’assurance”. UK IDNHR meeting. Biotechnical slope protection and erosion control. C. K. C. Flood Hazard Research Project. C. E . 1999. Center for Risk Management. Forestry Commission 1993. Fort Belvoir VA: US Army Corps of Engineers. “Runoff (floodwater) farming and rural water supply in arid lands”. Institute for Water Resources Green. Flood Control in the World: A Global Review. R. I. 1998. Westport. H. Floods . H. “Flood hazard management” in Herschy R W and Fairbridge R W (eds. “Making women visible in Disasters: Problematising the Private Domain”. Gilbertson. Denver CO: Bureau of Reclamation. D. T. M. Paris: OECD Gaschen. Menzinger. “ Men must work and women must weep” in Enarson. J. 1986.D. “Flood control measures in the River Thames catchment” in Rossi G. B. processes and practices. C.. “Sustainable development for river catchments”. Hausmann.swissre. report for the Hungary Flood Study. Vulnerability analysis for flooding: nodes and networks. and Silverman. published jointly by the USDA. C. W. 2000. 1998. Stream Corridor Restoration: principles. C.water. 4-5 Green. Natural Hazard Mitigation: Recasting Disaster Policy and Planning. Journal of the Institution of Water Engineers and Scientists 40 (3). C. 1998. T. (vol 8): 3: pp 308-320. B. 199.com) Gateley. Berke... DSO-99-06. report to Habitat/United Nations Environmental Programme. Dam Safety Office Gray. New Delhi: ICID Gardiner. 1996.1999.9 A Procedure for Estimating Loss of Life Caused by Dam Failure.an insurable risk. W. and Leiser. J. D. 229-248 This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. A. D.fdr) Fordham. H. Schaad. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . Journal of Institution of Environmental and Water Managers. J. The financial and institutional arrangements for flood defence in England and Wales. Framji.) Encyclopedia of Hydrology and Water Resources. Dordrecht: Kluwer Green. A. and Ketteridge.S. P. (eds.. conclusions. Disasters 22(2):126-143. P. C.an insurable risk? A market survey. 1976. Hausmann.) Through Women’s Eyes: the Gendered Terrain of Disasters. 1994. in SCOR Are natural hazards insurable?. Dordrecht: Kluwer Gardiner.gov/FEATReport120.. Zurich: Swiss Re (http://www. 1998. "Evaluating the intangible benefits and costs of a flood alleviation proposal". and Garg. 1999. K. Ct: Praeger..World Commission on Dams Thematic Study of Flooding 128 Federal Interagency Stream Restoration Working Group 1998. Discussion Paper 92-02. Applied Geography 6. E. Green. 1973. Final Report of the Flood Emergency Action Team (http://rubicon. 1994.. Enfield: Flood Hazard Research Centre Green. J. New York: Van Nostrand Reinhold Greeley-Polhemus Group 1992. A. Brower. TVA. 5-11 Glickman. and Kaiser. FEMA and US Army Corps Flood Emergency Action Team 1998. M. Forest and Water Guidelines. S. P. E. D. Agricultural Water Pricing in OECD Countries. T. Himganga 1(3). London. H..L. ENV/EPOC/GEEI(98)11/Final. 1999. D C: Resources for the Future Godschalk. Beatley. Special Publication No.
R. 1998. C. 18(6). Gueri. 27-30 Green.D. H. M. Enfield: Middlesex Polytechnic. C. Floods . Enfield: Flood Hazard Research Centre Green. R. Emery. W. and Horton. M. Flood Warning: Issues and Practice in the Total System Design. C.. and Emery. D. 482-511 Handmer. H. J.. 1998. Report to the Japan Institute of Construction Engineering. 1986. London: Middlesex University Press Green.J. Working Paper No. This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. British Columbia”. Vol. Tunstall.S. unpublished DEA thesis. and Helvey. and Morin. H. and Warner.. M. "Flooding and the Quantification of 'intangibles'". and Parker. D. 353-355 Herschy. H. V. P. and Penning-Rowsell. 221-230 Handmer. “The risks from flooding: which risk and whose perception”.) Encyclopedia of Hydrology and Water Resources. The views. F. “Health Hazards of Floods: Hospital Admissions for Lismore”. R. G. 1999. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . Disasters. The real cost of flooding to households: intangible costs. What People did during the Big Thompson Flood. W. Parker. River Thames Development Since 1800. Swindon: Halcrow Halcrow/VITUKI/FHRC/EXTERNAL/FOMI/MTA/Koros Valley District Water Authority 1999. “A Typology of Resilience: Rethinking Institutions for Sustainable Development”. conclusions..H. R. ‘Resilience. C. P.. Hausman.. W. Wierstra. Boulder.) Floods Across Europe. 227-236 Green.swissre. Avon. Health effects of Flooding: A Survey at Uphill. Flexibility and Adaptive Management . (eds. L. 10 (2).13 (2). and Fordham. E. Dordrecht: Kluwer Heurta.consecol.32. D. paper given at the Stockholm Water Symposium Green. and Penning-Rowsell. C. Hungary flood control development and rehabilitation project. “ Effects of forest clear-felling on storm hydrographs” Water Resources Research. J. P. D. 1996.) Natural Disasters: Protecting Vulnerable Communities. and Johnson. Penning-Rowsell. R. H. 1985.6. and Smith. London: Thomas Telford Green. 1989. W. I. Parker. 1970.. S. Nicholls. M. D. Gresent. Final report Handmer. and Fordham. C.. A framework for analysis and decision-making in the face of risks and uncertainties. P.L. H. 1991. (eds. Canadian Journal of Forest Research. Enfield: Flood Hazard Research Centre. J. Disasters 15(3). “The effect of slash and burning on water repellency of forest soils at Vancouver. 1987. “The Effects of the Floods Caused by El Nino on Health”. R. 1983. 2000. 541-546. Gonzales. C. “Designing for Failure” in Merriman P A and Browitt C W A (eds.org) Halcrow/FHRC 1996. 1978 “ Coping behaviour of elderly flood victims”. “Vulnerability Refined: Analysing Full Flood Impacts” in Penning-Rowsell. Washington DC: Resources and Technology Division. 1983. report to the Environment Agency. Vol. (ed. 118-24 Gunderson. National Policy of ‘No Net Loss’ of Wetlands: What Do Agricultural Economists Have to Contribute?. A.an insurable risk?. C. 768-82. E. 1993.World Commission on Dams Thematic Study of Flooding 129 Green. J.. “Dams: Failure” in Herschy. E. E. 1991. C. and Dovers. C. 1977. S. J. Essai d’evaluation de la politique publique de prevention du risque inondation: les cas des Vals d’Orleans et de Tours. Hewlett. van den Veen. 1994. Zurich: Swiss Re (http://www.com) Heimlich. Colorado: University of Colorado.J. 1999. 1996. E. F. Industrial & Environmental Crisis Quarterly 9(4). C. Paris: Universite Paris XII/ENGREF/ENPC Gruntfest.J. C. R. Enfield: Middlesex Polytechnic Green. “Flood Hazard Management: an emergent systems perspective”. and Penning-Rowsell. Natural Hazards Research and Applications Information Center.H. J.J. J. and Fairbridge. E.Antidotes for Spurious Certitude?’ Conservation Ecology 3 (1) (http://www. Journal of the Institution of Water and Environmental Management 3 (1)..D. Gerontologist.C. E. C.. and Golding. US Department of Agriculture Henderson. E. C. Australian Geographical Studies 21..) 1997. W.
. in Gruntfest. The total Flood Warning System: Concept and Practice”. New York: Academic Press Leach. 1997. D.) Floods London: Routledge. Cramer. Research.A. School of Civil Engineering. K. 527-43.M. 11(2): 207-45. Hughes. 1-23. International Joint Commission (http://www.32(4). (ed.W. University of New South Wales Hillman. “The Ecological role and Impact of floods”.J.org/boards/rrb) Islam. 1987.) Flood Warning: Issues and Practice in the Total System Design.) 1997. Lazarus. and Gilman.143. unpublished PhD. “Do land use changes in the Himalayas affect downstream flooding? Traditional understanding and new evidence”. C. R. and Wetmore.. World Disasters Report 1998.J.) Floods.A. The impacts of flooding and methods of assessment in urban areas of Bangladesh. Hong Qingyu and Luo Zhang-Yu (eds. (ed) Floods London: Routledge. Twenty Years Later: What We Have Learnt Since the Big Thompson Flood. T.ijc. Water Resources.A. Laska. 1996.) 1999. 959-74. Oxford: Pergamon.R. Adaptive environmental assessment and management.A. Jones. K. J.. K. In Parker.R. An Economic Comparison of Different Flood Mitigation Strategies in Australia. “Forest treatment effects on water yield”. and Simonivic. Psychological stress and the coping process.org/boards/rrb) Keys.S. Final Report – The Next Flood: Getting Prepared. Progress in Physical Geography. Delhi: International Commission on Irrigation and Drainage ICOLD 1999. L. unpublished PhD Thesis. World Register of Dams 98. J. 1-3 This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. J. G. E. J. in Parker.and Grant.1998. S. Guidelines for Forward-Looking Flood Protection. I. Juliano. 1993. S. Higgin. Chichester: John Wiley Holling. Hofer. “National Weather Service Advanced Capabilities in Flash Flood Forecasting”. J. I. Winnepeg: University of Manitoba (http://www. 1999. Institutional framework for water management in Hungary: EUROWATER-CEC vertical report.13.) Forest Hydrology. Berlin: Geschaftsstelle der Landerarbeitsgemeinschaft Wasser Larson. 1997. Oxford: Oxford University Press Ijjas. Oregon”. In Sopper. Huisman. S. 1997. ‘ Resilience and stability of ecological systems’ Annual review of ecology and systematics 4. and Sconnes. C. Vol21(4) 501-29. W. (ed. J. in Parker. D. G. The views.E. C. 2000. Special Publication No. R. 1973. Vol. No. T. 1967. 1997. The impact of wetlands on flood control in the Red River valley of Manitoba. (ed. Manual on Non-Structural Approaches to Flood Management. Flood Control and 1998 Flood of the Yangtze River (in Chinese) Hudson. A. Holling. Boulder. “Human Modification of Flood-Producing Processes: the Evidence from Catchment Studies.W. Delft: Netherlands Hydrological Society ICID 1999. J.A. R. Mearns. W. Jones. “Long-term variability in the water balances of the Plynlimon catchments”. and Szlavik. Memoir Geological Society of India. (ed. F. P. in Handmer. and Lull. 1981. “Peak flow response to clearcutting and roads in large and small basins. C. Landerarbeitsgemeinschaft Wasser 1995. “Reducing Flood Loss Potential by Flood proofing in the USA”. 1978. “Editorial: Community-Based Sustainable Development Consensus or Conflict?” IDS Bulletin 28(4).. L. and Hooghart. Water in the Netherlands.A. 1998. conclusions. Colorado: University of Colorado.ijc. 1997. 2000. Vol. 1966.E.World Commission on Dams Thematic Study of Flooding 130 Hibbert. S. Enfield: Flood Hazard Research Centre. J. 2000. “The initiation of natural drainage networks” Progress in Physical Geography. 119-141. Natural Hazards Research and Applications Information Center. van Ee. D. F. M. London: Routledge. (eds.355-80. M. Budapest: National Water Authority of Hungary International Red River Basin Task Force 2000. N. Paris: ICOLD IFRCRCS (International Federation of Red Cross and Red Crescent Societies) 1998. Enfield: Middlesex University Jones. H. Journal of Hydrology. “Floodplain biogeomorphology”. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . (eds.41 .) nd. Western Cascades.
London: Routiedge McMahon. 369379. “The general theory of second best”. 2000. V. A. Haines. Waterlogged Wealth. Flood and Coastal Defence: Project Appraisal Guidance Notes. 11-32 Ludwig. 1980. Leiden Nature Conservancy Council 1983. PWD 86010. “Global Change. Nisbet.. DC: Joseph Henry Press Ministry of Transport. London: Zed Books Obled. London: Earthscan Marsh 1988 “Thames. M. Ministry of Water Conservancy. and Yuming. 1999. V. J. Chinese Economic Performance in the Long Run. A.. O. H. . The views. 1990. London: HMSO MAFF (Ministry of Agriculture. Fisheries and Food) 1993. 1989. Michaels. London: Routledge Nicholls. E.utokyo. Government of New South Wales Newson.. 1992. Global Runoff Continental Comparisons of Annual Flows and Peak Discharges.. A. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . N.. Washington. D. 1990. and Tourasse. Water and Development. Gunaratnam. 1994. F. C. Floodplain Development Manual. D.World Commission on Dams Thematic Study of Flooding 131 Linnerooth-Bayer. and Yevjevich. H. “Coastal Megacities and Climate Change”. conclusions..S. J. London: Ministry of Agriculture. J. G. K. C.) Coping with Floods. 203-219 Lipsey. Lelystad: RIZA Mirtskhoulava. P. F. 2nd edition. A. Project Appraisal Guidance: 3 Economic Appraisal.W. J. G.. 1. Ericksen. and Yevjevich. J. A. Nature conservation and river engineering. Peterborough: NCC New South Wales Government 1986. Land. J . Dixon.html) Owen. “Environmental impact assessment for Xiaolangdi Yellow River multi-purpose economic-cum-enviromental improvement project”. “Uncertainty in flood forecasting: A French case study” in Rossi. Oi. Fisheries and Food) 1999.jp/Newsletters/2. Dordrecht: Kluwer Morris. D. 1995.. A. N. Les couts entraines par les inondations des terres agricoles en Ile de France estimation approximative. T. C. D. J. J. Paris: OECD MAFF (Ministry of Agriculture. Cremlingen-Destedt: Catena Mileti. 3.ac. “Flood control in the former USSR” in Rossi. Environmental Management and Governance: Intergovernmental Approaches to Hazards and Sustainability. Forestry Commission Bulletin86. INCEDE Newsletter 2(2) (http://incede. 37. Flood Emergency Plans: Guidelines for Corps Dams. J. London: Routledge Namafe.a community based approach. Review of Economic Studies 24. (eds. G. Disaster Mitigation. paper given at The People’s Role in Wetland Management Conference. Public Works and Water Management 1996. 1997. Munich: Munich Re Myers. 1996. A Survey of city's flood control in China. N. Chapter 14. F. Forest and surface water acidification. “Floods: Friends or Enemies?”. E. and Lancaster. Landscape Planning of the river Rhine in the Netherlands. Silsoe: Silsoe College Munich Re 1997. 1986.) Encyclopedia of Hydrology and Water Resources.iis. 1995.. Disasters by Design: A Reassessment of Natural Hazards in the United States. Research Document No. 1956-57. Beijing City: National Flood Control and Anti-Drought Headquarters. 1994. Burby. The Geneva Papers on Risk and Insurance 25(2). (eds. M.) Coping with Floods. S. “Nepal flood disaster. 1995. The Environmentalist 15. Dordrecht: Kluwer Maskrey. 2000. in Parker. E. 45-57 Maddison. Natural Disasters and Loss-sharing: Issues of Efficiency and Equity”. and Amendola. M-F and Passerini. and Rahnema. Nomani. 1994. and Slater. Fisheries and Food Maltby. Dordrecht: Kluwer Office of State Flood Control and Drought Relief Headquarters. Davis CA: The Hydrologic Engineering Center. E. R. Smith. Finlayson. 1989. Geojournal. B. London: Institute of Education Namafe. “The Dutch Hydro-Agricultural Rice Project in Western Zambia: An Assessment”. T. Islamic Economic Systems.. Handmer.R. P. 13. Flooding and insurance. River” in Herschy R W and Fairbridge R W (eds. T. 1994. R. and Srikanthan. Harmancioglu. Harmancioglu. L. 1993. Oxford: Oxfam May..2/2-2-3.) Floods. 1998. London:HMSO. S. US Army Corps of Engineers This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. R. “Floodplain management: historic trends and options for the future”. T. R. July 1993”. (ed.
World Commission on Dams Thematic Study of Flooding
Parker, D. J., Green, C. H. and Thompson, P. M. 1987. Urban Flood Protection Benefits: a project appraisal guide, Aldershot: Gower Parker, D. J. and Thompson, P. M. 1991. “Floods and tropical storms”, In The Challenge of African Disasters, 38-59, Addis Abbaba: World Health Organisation PanAfrican Centre for Emergency Preparedness and Response. Parker, D. J. 1995. “Floods in cities: increasing exposure and rising impact potential”, Built Environment, Vol.21, No2/3, 114-25. Parker, D. J. 1996. “International Experiences with Flood Hazard Management: Key trends and General Lessons”, paper presented at 6th BRAND 96 European Congress, 25 April, Amsterdam Parker, D.J. and Tapsell, S.M. 1995. “Hazard transformation and hazard management issues in the London Megacity”, Geojournal, (37) 3: pp 313-28. Parker, D.J. 1991. The Damage-Reducing Effects of Flood Warnings Report prepared for National Rivers Authority (Anglian Region) Regional Telemetry Scheme Appraisal, Enfield: Flood Hazard Research Centre. Parker, D.J. 1995. “Floods in cities: increasing exposure and rising impact potential”, Built Environment 21(2/3), 114-25. Parker, D.J. 1999. “Flood Warning Dissemination and Response: Lessons from International Research” in Proceedings of the 34th MAFF Conference of River and Coastal Engineers, Keele University, 30 June - 2 July, 1999. Parker, D.J. 2000. “Managing Flood Hazards and Disasters: International Lessons, Directions and Future Challenges”, in Parker, D.J. (ed.) Floods London: Routledge. Parker, D.J. and Handmer, J. 1998. “The role of Unofficial Flood Warning Systems”, in Journal of Contingencies and Crisis Management, 6(1): 45-60. Paul, B. M. 1984. "Perception of and agricultural adjustment to floods in Jamuna Floodplain, Bangladesh", Human Ecology 12(1), 3-19 Peerbolte, E. B. 1994. Sea-level Rise and Safety, Enschede: Unniversiteit Twente Penning-Rowsell, E. C. and Fordham, M. (eds.) 1994. Floods Across Europe, London: Middlesex University Press Penning-Rowsell, E. C., Green, C. H., Thompson, P. M., Coker, A. C., Tunstall, S. M., Richards, C. and Parker, D. J. 1992 The economics of coastal management: A manual of assessment techniques, London: Belhaven Penning-Rowsell, E. C., Winchester, P. and Gardiner, J. L. 1998. “New Approaches to Sustainable Hazard Management for Venice”, The Geographical Journal 164(1), 1-18 Penning-Rowsell, E.C., Winchester, P, and Bossman-Agrey, P. 1987. Flood Alleviation for Maidenhead: an evaluation of three levels of non-structural approach. Centre Internal Publication. Enfield: Middlesex Polytechnic. Penning-Rowsell, E. C. and Chatterton, J. B. 1977. The benefits of flood alleviation, Aldershot UK: Gower Technical Press. Penning-Rowsell, E.C. and Smith, D.I. 1987. Self-help flood hazard mitigation: the economics of house raising in Lismore, N.S.W., Australia. Tijdschrift voor economische en sociale geografie, Vol.78 (3), 176-189. Penning-Rowsell, E.C., Tunstall, S.M., Tapsell, S.M. and Parker, D.J. 2000. “ The Benefits of Flood Warnings: Real but Elusive, and Politically Significant”, Journal of the Institution of Water and Environmental Management, Vol.14, 7-14. Petts, G.E. 1984. Impounded rivers: perspectives for ecological management, Chichester: Wiley. Pielke, R. A. 1999. “Who Decides? Forecasts and Responsibilities in the 1997 Red River Flood”, Applied Behavioral Science Review (in press). Pottier, N. 1998. L’utilisation des outils juridiques de prevention des risques d’inondation: evaluation des effets sur l’homme et l’occupation des sols dans les plaines alluviales (application a la Saone et a la Marne), unpublished PhD thesis, Paris: Ecole Nationale des Ponts et Chausees Powell, B. and Penick, E. 1983. “Psychological Distress following a Natural Disaster: a one-year follow up of 98 flood vistims”, Journal of Community Psychology, II, 267-276.
This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. The
views, conclusions, and recommendations contained in the working paper are not to be taken to represent the views of the Commission
World Commission on Dams Thematic Study of Flooding
Purseglove, J. 1988. Taming the Flood, Oxford: OUP Ras Gharib, Hurghada, Safaga, Quseir, Marsa Alam, Shalatein, Abu Ramad and Halaib 1997. Hazard Assessment and Mitigative Measures of Flash Flooding on the Red Sea Towns, Egypt, NARSS Special Publication No. 1, Cairo: National Authority for Remote Sensing and Space Sciences (in Arabic) Rees, J. A. 1998. Regulation and Private Participation in the Water and Sanitation Sector, TAC Background Papers No. 1, Stockholm: Global Water Partnership Reiter, P. 2000. International methods of Risk Analysis, Damage Evaluation and Social Impact Studies concerning Dam-Break Accidents, Helsinki: PR Water Consulting Reppert, R. 1992. National Wetland Mitigation Banking Study, IWR Report 92-WMB-1, Alexandria VA: Institute for Water Resources, US Army Corps of Engineers Resolve Inc 2000. Participation, Negotiation and Conflict Management in Large Dams Projects, Thematic Review V.5, Cape Town: World Commission on Dams (http://www.dams.org) Rivers Bureau 1985. Rivers in Japan, Tokyo: Ministry of Construction Rivers Bureau 2000. Rivers in Japan, Tokyo: Rivers Bureau, Ministry of Construction (http://www.moc.go.jp/river/english/bureau/p16.html) Robinson, M. 1990. Impact of improved land drainage on river flows, Wallingford, Institute of Hydrology Report 113. Rogers, P., Lydon, P. and Seckler, D. 1989. Eastern Waters Study: Strategies to manage flood and drought in the Ganges-Brahmaputra Basin, Washington DC: USAID Royal Society for the Protection of Birds 1994. The new rivers and wildlife handbook, Sandy: RSPB Rutrough, J. E. 1997. “ Funding major disasters with traditional insurance” in Britton, N. R. and Oliver, J. (eds.) Financial Risk Management for Natural Catastrophes, Proceedings of a conference sponsored by Aon Group Australia Ltd, Brisbane: Griffith University Saad, M. B. A. 1999. “The Role of High Aswan Dam towards Nile Flood Control”, paper given at the Workshop on Benefits of and Concerns about Dams, Anatalya SANDRP (South Asia Network on Dams, Rivers and People) 2000. Comments received on the draft of this report SAST (Scientific Assessment and Strategy Team) 1994. Science for Floodplain Management, Washington DC: SAST (http://edcwww2.cr.usgs.gov/www/sast_aut.htm) Schmuck-Widmann, H. 1996. Living with the flood: survival strategies of Char-Dwellers in Bangladesh, Berlin: ASA Schware, R. 1982. “ Official and Folk Flood Warning Systems: an Assessment”, Environmental Management 6, 209-216. Sckade, D. A. and Feather, T. D. 1996. Environmental Valuation: the role of stakeholder communication and collaborative planning, IWR Report 96-R-17, Alexandria VA: Water Resources Support Center, US Army Corps of Engineers Shaw, 1998. “Floods: Largest in the USA, China and the World” in Herschy, R. W. and Fairbridge, R. W. (eds.) Encyclopedia of Hydrology and Water Resources, Dordrecht: Kluwer Shaw, B. D. 1984. “Water and society in the ancient Maghrib: technology, property and development”, Antiquites africaines 20, 121-173 Sheaffer, J.R. 1967. Introduction to Flood-proofing, Washington, D.C: US Army Corps of Engineers. Shultz, S. D. 1999. The Feasibility of Wetland Restoration to Reduce Flooding in the Red River Valley: A case Study of the Maple (ND) and Wild Rice (MN) Watersheds, Agricultural Economics Report No 432,, Fargo ND: North Dakota State University (http://www.ijc.org/boards/rrb) Sikander, A.S. 1983. “Floods and families in Pakistan - a Survey”, Disasters, 7 (2), 101-106. Smith D I 1990 “The worthwhileness of dam failure mitigation: an Australian example”, Applied Geography 10, 5-19 Smith, D. I. 1998. Water in Australia, Oxford: Oxford University Press Smith, D. I. and Handmer, J. W. (eds.) 1986. Flood Warning in Australia: Policies, Institutions and Technologies, Canberra: CRES, Australian National University
This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. The
views, conclusions, and recommendations contained in the working paper are not to be taken to represent the views of the Commission
World Commission on Dams Thematic Study of Flooding
Sorensen, J. H. and Rogers, G. O. 1989. Protective Actions for Extremely Toxic Chemical Accidents, Oak Ridge National Laboratory Sorensen, J.H. 1987. “ Evacuation due to off-site releases from chemical accidents: experience from 1980-1984”, Journal of Hazardous Materials 14, 247-257. Sorensen, J.H. 1988. Evaluation of Warning and Protective Action Implementation Times for Chemical Weapons Accidents. ORNL/TM-10437, Oak Ridge: Oak Ridge National Laboratory. Southworth, F. and Chin, S-M. 1987. “Network evacuation modelling for flooding as a result of dam failure”, Environment and Planning A 19, 1543-1558 Summer, S.G. 1997. “Comparison of Deficiencies Associated with the Big Thompson Flash Flood Event and Recent Flood Events in the Eastern United States”, in Gruntfest, E. (ed.) Twenty Years Later: What We Have Learnt Since the Big Thompson Flood, Special Publication No.13, Natural Hazards Research and Applications Information Center, Boulder, Colorado: University of Colorado. Swiss Re 1996. Insurance derivatives and securitization: New hedging perspectives for the US catastrophe insurance market?, sigma no. 5/1996, Zurich: Swiss Re Swiss Re 2000. Sigma 2/200 Natural Catastrophes and Man-Made Disasters in 1999 (http://www.swissre.com) Sydney Water Board 1985. Warragamba Dam Flood Protection Program, Sydney: Sydney Water Board Tapsell, S.M., Tunstall, S.M., Penning-Rowsell, E.C. and Handmer, J.W. 1999. The Health Effects of the Easter 1998 Flooding in Banbury and Kidlington, Report to the Environment Agency, Thames Region, Enfield: Middlesex University Flood Hazard Research Centre. Thomas, D. H. and Adams, W. M.1999. “Adapting to Dams: Agrarian Change Downstream of the Tiga Dam, Northern Nigeria”, World Development 27(6), 919-935 Thompson, P.M. and Sultana, P. 2000. “Shortcomings of Flood Embankment Strategies in Bangladesh” in Parker, D. J.(ed.) Floods, London: Routledge. Tobin, G.A. and Montz, B.E. 1997 Natural Hazards: Explanation and Integration, New York, N.Y: The Guilford Press. Torterotot, J-P. 1992. Analyse des reponse a l’annonce de crue et impact sur les dommages dus aux inondations dans plusieurs region de France, Noisy-le-Grand: CERGRENE Torterotot, J-P. 1993. Review of best practise in European strategic land use planning: French contribution, NRA R and D Project 426, Noisy-le-Grand: CERGRENE Trimble, S.W. Weirich, F.H. and Hoag, B.L. 1987. “ Reafforestation and the Reduction of Water Yield on the Southern Piedmont since Circa 1940” , Water Resources Research, 23:425-37. Tunstall, S.M., Tapsell, S.M. and Fordham, M. 1994. Public Perception of Rivers and Flood defence: Final Report, Enfield: Flood Hazard Research Centre. U.S. Army Corps of Engineers 1984. Flood Proofing Systems and Techniques, Washington, D.C. U.S. Army Corps of Engineers 1996. Flood Proofing Techniques, Programs and References, Washington, D.C. United Nations, Department of Humanitarian Affairs 1997. Floods: People at risk,strategies for prevention, Geneva: UN., United Nations 1992. Agenda 2 1, New York: UN United States Army Corps of Engineering 1995. Floodplain Management Assessment of the Upper Mississippi River and Lower Missouri Rivers and Tributaries United States Army Corps of Engineers 1998a. Flood Proofing Performance – Successes & Failures, National Flood Proofing Committee (http://www/usace.army.mil/inet/functions/cw/cecwp/nfpc.htm) United States Army Corps of Engineers 1999. US Army Corps of Engineers Annual Flood Damage Report to Congress for fiscal year 1998 (http://www.usace.army.mil/inet/functions/cw/cecwe/flood98/) United States Army Corps of Engineers nd. The Great Flood of 1993 Post-flood Report (http://www.mvr.usace.army.mil/PublicAff..e/HistoricArchives/Floodof1993/fl-l.htm)
views, conclusions, and recommendations contained in the working paper are not to be taken to represent the views of the Commission
1997. United States Federal Emergency Management Agency 1997. Human Adjustment to Floods. C. 1997. J. Atlanta GA: Centers for Disease Control and Prevention (http://www.R. Washington DC: United States General Accounting Office US Army Corps of Engineers 1998b. US Army Corps of Engineers Annual Flood Damage Report to Congress for fiscal year 1997 (http://www. and Spiegel. 1998. United States Committee on Large Dams 2000. R . 1958. C. Special Publication No. 93. 1945. New York: McGraw-Hill Walters.13.html) United States Federal Emergency Management Agency 1986. Washington DC: FEMA. M. G. and Control of Levees. Uncertainty in flood damage assessment: when does it matter? in Casale. W. P. P. C.. (http://www. Homeowner's Guide to Retrofitting.gov/nccdphp) Vazquez. Chicago: Department of Geography. Chicago: University of Chicago. and Volk.usace. Luxembourg: Office for Official Publications of the European Communities Wittfogel. Changes in urban occupance of floodplains in the United States. L. National Land Agency White. and Maza. London: Methuen Walters. 1967. “Challenges in adaptive management of riparian and coastal ecosystems” Conservation Ecology 1 (2) (http://www. conclusions. Water Management in Hungary at the turn of the Millennium. and Green. F. 1957. G. Budapest: Ministry for Transport. United States Army Corps of Engineers 1996. Water Resources in Japan.. Kok. London: MacMillan Water Resources Department 1996. “Geotechnical Reliability of Levees” in Burnham. and Samuels. The views.World Commission on Dams Thematic Study of Flooding 135 United States Army Corps of Engineers 1984. Shaeffer. Research Paper No. J. 1978.htm) US Department of Health and Human Services 1999. Department of Geography White. Dominguez.J. in Gruntfest. H.edu/hazards) Wagret. Quick Response Report #104. 1964. National Water Authority Waelde. Flood-proofing nonresidential structures. United States Federal Emergency Management Agency 1998. Koopman. C. Washington DC: USACE. Department of Geography White. B.. F. “Flash Floods in Mexico”. 1997. T. G. D.A. W. Chicago: University of Chicago.) . New York: Pergamon Wolff. Project Impact Guidebook: Building a Disaster Resistant Community. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . Choice of Adjustments to Floods. Polderlands. Midwest Flood: Information on the Performance. 29. Colorado: University of Colorado. Adaptive management of renewable resources.colorado. F.army. Washington DC: FEMA United States General Accounting Office 1995. 57. Pedroli. GAO/RCED-95-125. H. 1986.. G. management and flood mitigation.. B. Oriental despotism.) Ribamod: River basin modelling. W. Fuentes. J. 1997. Programs and References. Boulder.M. M.W... SP-28. Floods: A Geographical Perspective. Effects. Flood Proofing Techniques. D.) Hydrology & Hydraulics Workshop on Risk-Based Analysis for Flood Damage Reduction Studies. Mayer.cdc. R. D. Calef.C. Boulder CO: Natural Hazards Center (http://www. Ward. Washington DC: USACE. de Blois. Washington DC: FEMA. M. Vituki 1998.. Peerbolte.. Research Paper No. Natural Hazards Research and Applications Information Center.consecol. J.privatei.F. O. and Davis. Davis CA: US Army Corps of Engineers Hydrologic Engineering Center This is a working paper prepared for the World Commission on Dams as part of its information gathering activities. (eds. E. Tokyo: Minister’s Secretariat.. C.org). H. Dissociative and posttraumatic reactions to the Northern California flooding of 1997. Targeting Arthritis: The Nation’s Leading Cause of Disability. (ed. C. R. K. Communication and Water Management. Twenty Years Later: What We Have Learnt Since the Big Thompson Flood. Hudson. G. University of Chicago Wind. Flood Proofing Systems and Techniques.com/~uscold.mil/inet/functions/cw/cecwe/flood. (eds. Research Paper No.
Knoxville: Association of State Floodplain Managers Wright. 1990. 1992. Dordrecht: Kluwer. J.org //html/edi/sourcebook/sbhome. The World Bank Participation Sourcebook (www. Co: Natural Hazards Research and Applications Information Center World Bank 1996.) 2000. Special Publication No. “Floods and society”. In Rossi.26. M. N. L. and recommendations contained in the working paper are not to be taken to represent the views of the Commission . R. NATO ASI Series. This is a working paper prepared for the World Commission on Dams as part of its information gathering activities.B. Sendek. “An economic and historical perspective of flood damage: the viability of structural solutions”. D.World Commission on Dams Thematic Study of Flooding 136 Wolman.worldbank. Procs. 1657-67 Yevyevich.. Report No.K. G.. R. “Logging effects on streamflow at Caspar Creek in northwest California”. M. 3-9. Strengthening Local Flood Protection Programs. Geological Survey Paper 282-C. Project Appraisal Document for the Xiaolangdi Multipurpose Project Stage II.H. Rice. conclusions. Pittsburg. (eds. and Thomas. The Nation’s Responses to Flood Disasters: A Historical Account.H. of the 10th Annual Conference of the State Floodplain Managers. V.. 15. G. River flood plains: some observations on their formation. 1986. Washington DC: US Government Printing House Wooley. V.htm) World Bank 1997. and Yevyevich. B. and Leopold. 1957. Boulder. Vol. Harmancioglu. US Geological Survey. 16274-CHA Wright. June 17-19. (ed. The views.) Coping with floods. Water Resources Research. K.
This action might not be possible to undo. Are you sure you want to continue?
We've moved you to where you read on your other device.
Get the full title to continue listening from where you left off, or restart the preview.