FLOOD00113 - Laserfiche WebLink

Home Browse Search

294 J.LAVABRE ET AL. HYDROLOGICAL RESPONSE OF A MEDITERRANEAN BASIN 295 100 22 On the other hand, the changes in the flood regime concern the shape of the hydrographs. The post-fire responses recorded are sharper and quicker, becoming almost impulse responses. In fact, runoff does not require previous saturation of the basin, and therefore water immediately reaches the drainage network since the retention power is lost. However, more precise studies are required to assess the changes produced in mechanisms acting in the runoff generation at short time steps. These conclusions are related to the changes in the hydrological response of the basin 1 year after the fire. Of course, the hydrological consequences of fire,like any others, will evolve with time controlled by vegetation growth and erosion processes. Therefore this study can be taken as if it was the first 'photograph' of the basin a year after the fire. Comparison with complemen- tary 'photographs', taken during subsequent years, will be needed to describe the evolution of the hydrological consequences of fire with time. Thus this work should be a starting point for in-depth studies of more accurate flood descriptions (using models of flood generation), and also for comparing basins affected to different degrees and analysing the evolution of the response during subsequent years. These studies are currently in progress. TABLE 10 Different estimations of the increment of annual runoff in 1990 Annual runoff increment mm % linear regression of the annual runoff (eqn. (2)) Difference from the envelope deficit in the Vaubarnier basin (Fig. 3) Difference with the linear regression of the initial deficit (eqn. (3)) Difference with the annual runoff given by OR3 monthly simulation Difference with the annual runoff given by GR3 daily simulation 112 225 24 64 148 32 35 monthly and yearly time steps, is the reduction of the loss power or equi- valently the increment of the annual runoff yield. At the daily step, the increase of runoff production is strictly related to the high water days and especially to the peaks of flow. ' (2) As for the flood regime, the destruction of the vegetation cover and soil alteration by fire have more crucial consequences. The retention power of soil and vegetation is lost, and the hydrological mechanisms producing runoff are significantly changed. We stated that the frequency of floods is significantly altered. The 10 year return period flow was exceeded three times in the post-fire year by rainfall events which did not exceed the yearly return period of the 12 h rainfall. This evidence may lead to the suggestion that, at least, the change in fiood frequency is that it converts the 10 year return period ft.ow into a 'I year return period'. However, we should note how highly dubious these extrapolations are. Therefore we prefer to stress the extreme fragility of the usual methods of rare frequency ft.ood evaluation, which are based on extrapolations of data of a period of a few years (note that the probability of a fire or another important change in the basin conditions will be, of course, higher). However, an important repercussion can be stated. The results found here indicate that the hydrological consequences of fire will also affect the hydraulic works and the drainage networks (normally dimensioned to the 10 year return period flow), since the changes in the hydrological response would make the design frequ~ncy analysi~ obsolete. Furthermore, in areas of con- siderable fire risk it could be interesting to think how to take the eventuality of a fire into account in the design flood. ACKNOWLEDGEMENTS This study has been carried out mainly through financial support from the French Ministere de la Recherche Scientifique and from the Conseil Regional Provence-Alpes-Cote d'Azur. The authors have also been supported by a Personal Investigador Altament Qualificat contract from the Generalitat de Catalunya, financial aid from the Spanish Plan Nacional NAT89-1072-C06- 01 and by a BDI grant from the C.N.R.S. Thanks are also due to C. Michel, of the CEMAGREF, for allowing us to use the GR3 model and to R. Rycroft of the Servei d'assessorament Linguistic (Universitat de Barcelona) for the English corrections. REFERENCES' Anderson, H.W., 1949. Does burning increase surface runom J. For., 47: 54-57. Anderson, H.W., Hoover, M.D. and Reinhart, K.G., 1976. Forests and water: effects of forest management on floods, sedimentation and water supply. USDA For. Servo Rep., Gen. Tech. Rep., PSW.18, Berkeley, CA. -The CEMAGREF and the OIS.Real Collobrier reports are available on written request to Mr. Jacques Lavabre. Copies of the lnstitut National Poly technique de Grenoble (INPG) Ph.D. thesis are available on written request to Atelier National de reproduction des theses. B.P. 47X. F-38040 Grenoble, Cedex, France.