In the last decade Europe has experienced a number of unusually long-lasting rainfall events, resulting in severe floods in most European countries, being the devastating and costly floods in the Elbe and Danube rivers in August 2002, the Odra flood in 1997, the UK floods in 2000, and the Rhine/Meuse floods in 1993 and 1995. More recently the floods in the alpine region in August 2005 caused great damage. The trend for increased precipitation and flooding seems to be continuing, which calls for action from politicians, civil services, economy and science aimed at the preparation of measures for the protection against extreme events.

It is clear that emergency, civil and water management agencies benefit from an increase in lead-time to efficiently implement their plans in downstream areas to reduce the flood peak. Due to the inevitable uncertainties in the flood forecasting chain the assessment of these uncertainties can help a decision maker to decide what measures should be taken. Therefore forecasts should not only provide a best guess of the state of the river system, but also an estimate of the range of possible outcomes. Ensemble forecast techniques can be used to obtain this kind of information. They allow effects of a wide range of sources of uncertainty on hydrological forecasts to be accounted for.

Working with ensemble forecasts and dealing with uncertainties in flood forecasting is a rather new discipline. Until recently this technique was hardly suitable for operational flood forecasting. Due to an increase in computation power and data transmission rates we are now in a position to use ensemble weather predictions effectively. Many scientific questions need to be addressed however to use them to their full potential. Some of the problems to discuss are:

• How can ensemble weather forecasts help to improve hydrological forecasts?
• How can hydrological ensemble forecasts be verified, also for big events, and what can be done to gain confidence that a given forecast system is reliable?
• How should ensemble weather forecasts be used in operational mode (calculation time, selection of representative ensemble members)?
• How does the uncertainty in weather forecasts translate into hydrological uncertainty?
• How can uncertainties in hydrological models, model parameters and hydrological initial conditions be represented in hydrological ensemble prediction?
• How should uncertainty be communicated to decision makers and to the public?
• What is the role of a human forecaster?
• What processes and tools are needed for forecasters to control the operation of a hydrological ensemble forecast system?

The International Commission for the Hydrology of the Rhine Basin (CHR) and the World Meteorological Organization (WMO) invite scientists, operational forecasters and decision makers in the field of flood management to discuss these and other issues during a two day workshop.