Last Updated on Saturday, 24 April 2010 19:14 Written by Administrator Saturday, 24 April 2010 18:51
Most of the methods routinely used for flood frequency estimation are designed to: i) be conceptually simple and easy to use, ii) not require too much in terms of volume and formatting of input data, and iii) make predictions based on an underlying assumption of a stationary environment.
However, with the increase in data availability, advances in both hydrological and statistical modelling techniques, and a general increase in available computational power, the development of the next generation of methods for flood frequency estimation should not be bound by these constraints.
Two different types of methods are used in practice for flood frequency estimation, where the choice between the two is often dictated by factors such as: national or institutional tradition, modeller expertise, complexity and objective of study, legislative requirements, and data availability. For relatively routine applications where only an estimate of the peak flow value is needed, direct and regional statistical extreme value analysis of river flow data might be sufficient. If more detailed information is required, such as the evolution of a flood event in time (a hydrograph) then more complex methods based on coupled system representing both rainfall (stochastic weather generators) and the terrestrial runoff processes (rainfall-runoff models) can be used. The latter models can provide more detailed information on the hydrological system, but often require substantially more input data and a higher level of expertise by the user. The aforementioned approaches are usually calibrated within specific regions and often within administrative boundaries. The specific scientific challenge in this respect is the pan-European comparison and evaluation of the consistency of estimates across methods, physiographic regions and a variety of spatial scales in order to ensure comparable flood frequency estimates and safety measures over Europe.
The increased need for policy makers and flood risk managers to consider the impact of environmental change on future flood risk, as required by the European Directive on Assessment and Management of Floods (Directive 2007/60/EC, hereafter referred to as the European Flood Directive), call for the next generation modelling systems to provide estimates of the effects of these changes on flood frequency. It is of utmost importance for the implementation of the Flood Directive that state of the art and harmonized methods are used to estimate extreme flood frequencies, e.g. to obtain consistent values for locations where rivers cross national borders. In particular, the effect of a changing climate on the frequency and severity of extreme hydrological events, such as floods, require the scientific community to develop appropriate tools for policy makers. Results from global and regional large-scale climate modelling studies have indicated that due to climate change, Europe will have to cope with more frequent and more severe flood events in the future than under current climate conditions. The scientific challenge is to down-scale the results from the large-scale studies to more operational tools enabling policy makers to assess the impact of climate change on a local scale.