Evaluation of the most important factors effecting maximum flood discharge using sensitivity analysis of empirical formulae

One of the most important tasks in catchment hydrological analysis is flood and flooding, and therefore estimation of river peak discharges is necessary. Peak discharge estimation as design flood is the basic and necessary element of water related projects and designs. In ungaged catchments, where no enough measured data is available, empirical approaches are usually applied to estimate maximum flood discharge. They are usually based on one or more factors such as drainage area that causes flood, and most of these methods have been proposed for a certain area with specific physical and climatic conditions. To use these methods in other areas with different conditions, evaluation and analysis of sensitivity of their parameters seems to be necessary. In this research, a new simple and efficient method is used to carry on sensitivity analysis of 10 empirical flood estimation methods including Creager, Diken, Fuller, Fanning, Inglis, Coutange, Mayer, Horton, USGS and German, in which some of them are the most famous existing methods. The results show high sensitivity of all equations to area in its lower ranges. In other word, small changes in catchment area in small watersheds cause high variation in model output (peak discharge). Also in most of the equations, the role of area is decreased in comparison with C coefficient in large watersheds. It means that while the area of catchment increases, its role on peak discharge decreases. In some equations such as Fuller, the output is also very sensitive to return period especially in the lower range return periods.