Flood risk evaluation using stochastic, conceptual and empirical models

Abstract:

The evaluation of the risk of extreme flood was always an important problem for the installation of a country because of the concentration of the human activities and of great lanes in the valleys. Flood risk mitigation is usually based on design floods, which have a large magnitude when the potential damage is great. However, extreme flood estimation is by definition a difficult task as it usually concerns a limited number of such events during a human lifetime. The work aims to answers the questions of which decision makers always have needs for answers: Which is the frequency of the extreme events? Which is their return period? And which are the rains of projects and the flows to dimension the hydraulic works?

Hydrological models can be defined as physical, mathematical or empirical. The latter class uses mathematical equations independent of the physical processes involved in the hydrological system. The linear regression and Gradex (Gradient of Extreme values) are classic examples of empirical models. However, conventional empirical models are still used as a tool for hydrological analysis by probabilistic approaches. In many regions in the world, watersheds are not gauged. This is true even in developed countries where the gauging network has continued to decline as a result of the lack of human and financial resources. Indeed, the obvious lack of data in these watersheds makes it impossible to apply some basic empirical models for daily forecast. So we had to find a combination of rainfall-runoff models in which it would be possible to create our own data and use them to estimate the flow. The estimated design floods would be a good choice to illustrate the difficulties facing the hydrologist for the construction of a standard empirical model in basins where hydrological information is rare. The construction of the climate-hydrological model, which is based on frequency analysis, was established to estimate the design flood of some Moroccan catchments. The choice of using this complex model returns to its ability to be applied in watersheds where hydrological information is not sufficient. It was found that this method is a powerful tool for estimating the design flood of the watershed and also other hydrological elements (runoff, volumes of water, etc.).