Flood risk: prevention, preparation, protection

From research to rollout. Prevention and warning systems are supplied with acquired data and estimation methods in real time to create maps of flood plains, optimize warning systems, run diagnostics for hydraulic structures, etc. Irstea is an expert in all 3 Ps: prevention, preparation, protection.


Prevention comes before warnings. Which flood plains should be monitored? What measures can be implemented to avoid significant damage? Research data is used to support prevention policies.

Public policies. What resources can help prevent flood risks?

Identifying areas at risk, followed by priority areas and implementing management plans: Irstea is contributing to the implementation of public prevention policies.

What exactly is meant by prevention?

SHYREG, an important database

The SHYREG database uses an original method developed by Irstea to raise regional awareness of high water flows. This involves estimating hydrological hazards based on a combination of two models, a rain generator and a rainfall-runoff model, the parameters of which are adjusted regionally so they can be used anywhere in the country. For example, for the hydrological model, regionalized parameters include local characteristics such as land use, topography or the type of subsoil. This method also provides rainfall statistics for the entire region at a resolution of one square kilometer, as well as runoff statistics for the entire hydrographic network. This method is particularly interesting as it can provide information for ungauged rivers (rivers without measuring stations).

The result of this research is a national database that is available for use by the Ministry of Ecology and decentralized national services as part of the flood prevention process. In particular, the data is used to create hydrological risk maps, as required by the European Flood Directive.

ContactsPatrick ArnaudPhilippe CantetJean OdryCatherine Fouchier (Aix-en-Provence)


Forecasting tools are used to calculate river flows using data from observed (or predicted) rainfall, as well as to produce spatial or temporal warnings. This is operational research!

Flood risks: forecasting floods to improve warning systems

Methods and tools are being developed to support warning systems. This is real operational research!


Natural risks in mountain areas: a real-time warning system in the PACA region

Warning maps produced using new generation hydrometeorological radars, adapted to mountain observations.


Flood risks in urban areas: French expertise exported to Brazil

A new weather radar has been set up in São Paulo, Brazil. Coordinated by the French company NOVIMET in partnership with Irstea, the project aims to improve the management of flood hazards. Researchers have adapted their forecasting models to the climate and urban setting.


Managing risk with probable flood scenarios: it's possible!

Using probability to manage hydrometeorological risks such as floods and high water is a new approach to forecasting that is gaining popularity in operational hydrology. Irstea is looking into it.

Flow forecasting models are often fed by rainfall forecasts. Rainfall information is therefore transformed into flow data. This type of model also includes information relating to initial conditions in river basins. The aim is to follow the development of flows and water levels in a river into a future ranging from several hours to several days. If potentially dangerous situations are predicted, maps can be traced to highlight areas at risk and help authorities and local populations implement protective measures.

However, significant uncertainty remains throughout the forecasting sequence. Beyond a few days, or even a few hours depending on the predicted situation, a forecasting system must take uncertainty into account to ensure the quality of the predictions issued. Both probabilistic and overall forecasting can be used to take these uncertainties into account and improve the characterizations of risks. They offer forecasters a clearer vision of potential future scenarios and associated probabilities. Probable flood scenarios can therefore be created to provide decision-makers with better information and further refine warning systems.

In partnership with Schapi (French National Hydrometeorological and Flood Forecasting Center), Météo-France, EDF and several international partners, Irstea researchers have been working for several years to resolve the many challenges of overall hydrological forecasting.

Contact: Maria-Helena Ramos (Irstea Antony - HBAN), Research Associate and Coordinator of the HEPEX international network for overall hydrological forecasting.



Neglected or poorly maintained flood protection structures can no longer fulfill their protective role. This type of failure can have more severe consequences than the flooding which would have occurred naturally in their absence. For this reason, the design, monitoring and maintenance of dikes and dams is a central aspect of significant safety and socioeconomic issues.

Irstea has developed expertise linked to the safety of hydraulic developments and structures by developing diagnostic methods, models and tools as well as evaluating performance and analyzing risks linked to dikes and dams. The institute provides technical support to both central and decentralized services (DREAL) for the Ministry of Ecology as part of scientific and technical support measures to manage dams and dikes. It is also involved in several working groups to issue technical recommendations and other expert documentation.

Dikes and dams: a special report on Irstea's expertise