Scope of research

Back to top

During a flood event, the responsible authorities need to make decisions on operation of barriers and on evacuation/rescue strategies. Temporary flood protection is sometimes an option and if so it must be decided on quickly. Also people are served by knowing evacuation routes and a forecast of blocked roads. A decision support system (DSS) can assist the user by quickly showing the available information on water levels, objects at risk, evacuation routes, resident vulnerability, etc. This support can be given during the preparation of flood event management or during the actual flood event.

The specific objectives of task 19 were:

• To obtain insight into the different types of decision support systems in flood risk management that have been made in the past or that are currently being used, and determine which of these DSSs would be most suitable for flood event management planning;
• To develop a methodological framework that allows for integration of the information that different end-users require;
• To implement this methodology in two outline DSSs, and test them in two pilot sites.
• To show how detailed two-dimensional hydrodynamic model results can be used in the preparation of community safeguard plans for urban areas.

The work was structured as follows:

• A review was carried out on decision support systems in Europe, to get insight in previous experiences;
• Different end-users were consulted to learn about user requirements for flood event management;
• A methodological framework was developed based on this first activity, to be applied in the three pilots of the UK, the Netherlands and France;
• For two pilots a prototype DSS was prepared based on this methodological framework, and tested among the end users.
• For the French pilot sites a two-dimensional approach to the preparation of flood event management plans was validated;
• Overall conclusions were drawn as well as pilot specific conclusions and recommendations.

A strong link was found with the work carried out under task 17 (Lumbroso et al, 2007) on evacuation and traffic modelling. Part of the work has been reported in FLOODsite technical notes.

Detailed description of work can be found in :

• Work package description;
  
• Task Review (Grenoble, February 2008).

Flood event management

Back to top

Flood event management deals with the following tasks:

• Short-term actions to prevent flooding (operation of barriers and retention areas);
• Actions to reduce the impact of flooding (opening or closing gates in the hinterland);
• Preparation of evacuation plans;
• Preparation of rescue plans;
• Evacuation before or during a flood event;
• Rescue after a flood.

The content of the tasks slightly differs between different natural systems, such as lowland areas with fluvial or tidal flooding, sloping areas with fluvial flooding and mountainous areas where flash-floods can occur. All tasks aim to reduce the impact of flooding in terms of casualties and affected persons, thus economic consequences are not considered.

Flood event management merely takes place in the lower half of the disaster cycle (Figure 1), i.e. in the preparation and response phases. Within the preparation phase the management of the actual event is planned. For example evacuation plans are designed. The preparation phase also consists of measures that can be taken in the time between the forecast of a flood event and the actual event. Examples of these measures are operation of barriers and retention areas and temporary raising of dikes with sand bags. Long term flood prevention measures, such as lowering of the floodplain are not incorporated.

Figure 1: A disaster cycle with focus on flood event preparation and response

In the response phase the aim is to reduce the flood consequences. This can be done by influencing the way the flooding proceeds by opening or closing gates in the hinterland, or by evacuation and rescue. Evacuation deals with the relocation of humans, livestock and capital goods from an area threatened by flooding to a safe place. Evacuation takes place before roads are blocked or houses are being flooded. The evacuation activity is started during the preparation phase, when the forecast is given. Rescue takes place during the flood event. The amount of resources used for rescue and the way they are deployed needs to be determined beforehand. An important issue to be addressed in flash flood areas is the state of the transport network to be used by rescue services.

Link with long-term planning

Back to top

In the current research a DSS is a computer-based information system that supports the ability of authorities to make flood management decisions at the regional level. Whereas task 18 dealt with long-term planning issues in flood risk management, task 19 focused on the ‘operational’ management of an actual flood event. The term operational refers to the fact that the end-user must take immediate action without sufficient time to perform model runs. Examples of actions that require immediate initiation include the closure of dams and gates, deciding which areas are to be evacuated and which people will require rescue efforts.

Other differences between flood event management and long-term flood risk management are:

• Time horizon: In flood event management, the time horizon is much shorter than in long-term FRM.
• Management options: Long-term FRM decisions are taken on a higher level and management options comprise policy directions or strategic alternatives, not ready-to-implement measures (see also De Bruijn et al, 2008). In flood event management the options are much more detailed, location-specific and ready to be implemented.
• Types of impact: The planning of flood event management concerns people's deaths and not so much the economic damage, in contrast to long-term FRM, where both aspects are as relevant.

Links to other projects and FLOODsite activities

Back to top

The current study is related to other activities carried out in FLOODsite (see Figure 2):

• FLOODsite task 8
  There was interaction with task 8 on flood inundation modelling in the urban flooding pilot in France;
• FLOODsite task 10
  Interaction with the Thames pilot related to the socio-economic evaluation and modelling technologies;
• FLOODsite task 14
  Results of the flood risk analysis carried out for this task on ‘Long term strategies for flood risk management’ (De Bruijn et al, 2008) were used to develop evacuation strategies;
• FLOODsite task 17
  Results obtained in this task on Evacuation and traffic modelling (Lumbroso et al, 2008) have been used in the development of the DSSs;
• FLOODsite task 18
  The methodological framework was developed in close cooperation with task 18, which aims at decision support for long term flood risk management. The review of DSSs was also carried out jointly with this task (McGahey et al, in prep);
• FLOODsite tasks 24 and 25
  The results of the current study have been used in the workshops organised in task 24 and 25. Findings are reported in the book ‘Methodologies for Integrated Flood Risk Management – Research Advances at European Pilot Sites’ (Schanze, in prep).

Figure 2: Scheme showing the interaction and links with other task of the FLOODsite project

Review of existing DSSs

Back to top

As a first activity a review has been carried out on available decision support systems in the UK, the Netherlands and France (Maaten et al, 2007). Table 1 provides a summary of the systems that have been reviewed, their functionality and their current end users. The reviewed decision support systems are rather different, as not all of them were developed for the purpose of flood event management (FEM). Some systems were made for flood prevention planning and long term flood risk management (FRM).

Nevertheless there are a number of resemblances. All the described systems are more or less "generic". They may have been set up and applied for a specific area only, but their modular set-up would with some effort allow application for other areas as well. Also some form of Geographical Information System (GIS) is present in most systems. For some systems only in the form of a simple map, just for orientation purposes, but usually there are different layers to show spatial information on various subjects.

In DSSs where results of model calculations have to be taken into account in the decision making, those results are usually not calculated real-time. In most models a selection is made from pre-calculated sets of model output. Exceptions are FLIWAS, where in the evacuation module an adaptation of the traffic model results to the actual situation during an evacuation will be possible, and PACTES, where efforts are made to produce real-time run-off predictions on the basis of (forecasted) rainfall.

Some systems are basically a DSS, with the decision part for the emergency managers, but they contain also a public part (web-based) for providing up-to-date information to the public during an emergency, via internet. Surrey Alert and FLIWAS are examples of such systems.

User requirements

Back to top

The FLOODsite Technical Note on user requirements in flood event management (Logtmeijer, 2006) provides an overview of the description of user requirements as they have been gathered by studies in literature, contacts with end-users and interviews with stakeholders and experts.

Methodological framework

Back to top

Based on the user requirements a methodological framework for flood event management DSSs was developed (Figure 3). It shows the eight modules that are generally relevant in flood event management (Van der Vat et al, 2007):

• The external driver module describes the existing situation prior to the flood and the boundary conditions for the flood event;
• The tools module consists of the tools used in the other modules;
• The management response module describes the management options available to the decision maker;
• The boundary conditions of the flood event such as fluvial / tidal water levels from a flood forecasting system form the input of the hazard module. At the level of the hazard module, the decision maker can influence the flood characteristics by preventive actions such as operation of barriers and retention areas, and actions that reduce the impact such as placing sand bags or closing gates;
• The exposure module compares the information on the flood characteristics with information on the distribution of inhabitants, livestock, property and utilities. The exposure can be influenced by execution of an evacuation or by a rescue operation.
• The vulnerability module defines the potential for the receptors (e.g. people, livestock and buildings) to be harmed. The input to the vulnerability module is based on a series of empirical or theoretical damage functions for each receptor, providing a relation between flood characteristics and level of harm.
• In the consequence module a damage and casualties model combines the exposure and the vulnerability and calculates the damage to people, livestock, property and utilities. If management options such as evacuation have been selected, their influence will be incorporated through the resulting effect in the exposure.
• The risk module combines the results of the consequence module for the different breach locations. The combined risk is expressed as the expected damage of a forecasted flood event under the selected management option.

Figure 3: Summarized methodological framework for flood event management DSSs

The methodological framework was developed in close cooperation with task 18 on long-term flood risk management DSSs (McGahey et al, unpublished). There are small differences between both frameworks, as follows:

• In flood event management, the time horizon is much shorter than in long term FRM.
• The probability of the flood event, given an extreme water level, is much higher than the average flood probability (per year). Therefore, the consequence and the risk values in FEM are much closer.
• The exact value of the calculated risk is not that important in FEM, it is the difference with the risk under other management options that counts.
• In long term FRM decision are taken on a higher level and management options comprise policy directions or strategic alternatives, not ready-to-implement measures (see also De Bruijn et al, 2008). In flood event management the options are much more detailed, location-specific and ready to be implemented.

Pilot applications

Back to top

Three pilots were chosen to further develop the existing knowledge on flood event management decision support. In each pilots different components of the methodological framework were implemented, based on the user requirements of the pilots.

The DSSs for the Thames and Schelde pilot allow users to compare hazards and risks related to flood event management, using the common hydrodynamic model output as a basis. Various management options and their effect can then be tested. All aspects included in the DSSs are shown in Figure 4 in a ‘technological framework’.

The French pilot dealt with urban flash floods. The use of a detailed two-dimensional (2D) hydrodynamic model was investigated and the contribution of the results to community safeguard plans was assessed.

Figure 4: Technological framework of the DSS development for the Thames and Schelde Estuaries

top