Considering tributaries to reconstitute an extreme event on a portion of the Medjerda river

The flooding problem was been imposed to the man since ever. Many natural disasters have led him to reflect on the power and frequency of floods. In the past, interest was mainly the knowledge of the maximum values of the water depth and flow. Currently, interest is turning more and more to the characteristics of hydrographs and their variation during flood propagation of river's tributary. Models of flood propagation are varied; there are a big number of algorithms from simple statistical receipt to the partial differential equations of Saint-Venant [1]. To reconstitute flood's hydrographs, the simplest models used until now are Muskingum and Regression. In Tunisia, the flood problem arise the only perennial river, Medjerda, in particularly the plain of Bou Salem. During the work of my memory Masters, we applied Muskingum and Regression models to reconstitute flood hydrographs of the main stations Medjerda River upstream Sidi Salem dam. We concluded that the model more efficient is regression [2]. We will try in this article to analyze the effect of introducing the tributaries in the reconstitution of flood hydrographs. The objective of this study is the reconstitution of flood's hydrographs of January 2003 of Bou Salem which is a subwatershed of Medjerda River. Two models were used: Muskingum and regression. Reconstitution times are 2, 4, 6 and 8 hours. To evaluate the quality of reconstitution models, three performance criteria were selected: standard deviation error `S', peak relative error `S1' and peak time difference `S2'. The reconstitution of Bou Salem' floods (downstream station) from Jendouba (upstream station) was sets in four scenarios: (1) ignoring the tributaries, (2) considering the tributary Tessa, (3) taking into account the tributary Mellegue and (4) considering both tributaries of Tessa and Mellegue. Following this application, we concluded that regression model reconstituted well the flood's hydrographs of January 2003 for all calculation times. Consideration of tributaries improves the results of reconstitution. The performed model obtained is the regression that takes into account two tributaries of Tessa and Mellegue at once, and for all calculation times. The model gave the best value of the performance criteria.