Apport du recueil de données historiques pour l’étude des crues extrêmes de petits cours d’eau. Etude du cas de quatre bassins versants affluents de l’Aude

Usefulness of historical flood series reconstruction, for the study of extreme floods in small watersheds. Case study of four small tributaries of the Aude river, France

¹ DDE du Gard – Service de Prévision des Crues du Grand Delta, 89 rue Weber 30907 Nîmes cedex
² Ecole Nationale des Ponts et Chaussées – CEREVE, 6-8 avenue Blaise Pascal, Cité Descartes, 77455 Marne-la-Vallée Cedex 2
³ Laboratoire Central des Ponts et Chaussées – Division Eau, BP 4129, 44341 Bouguenais Cedex

Auteur de correspondance : gaume@cereve.enpc.fr

Abstract

The results of an investigation on historical floods in small watersheds of the Aude region in France are presented herein. Historical information was gathered using archive documents, mainly located in local archive founds. Using this information, the discharges of the main historical floods could finally be estimated for a period ranging from one to two centuries, which highlights the feasibility of historical series reconstruction, even in the case of small rivers. The use of this data, in addition to systematic discharge measurements (continuous series of about 30 years in each case) shows that the calibration of Gumbel and Frechet distributions is highly modified using historical data. Moreover, the calibration obtained using historical data is much more accurate, despite of the poor accuracy of discharge estimations for historical floods. In some cases, it is also possible, with a simple test using historical data, to exclude a statistical distribution that is not adapted to represent both historical and systematic data samples. Lastly, the Gradex method is also tested and gives results in good coherence with historical data, despite of the high variability of flood discharges among the four studied watersheds.