Calibration of channel roughness in intermittent rivers using HEC-RAS model: case of Sarimsakli creek, Turkey

Natural channel's roughness determination constitutes one the most challenging procedure towards the development of hydraulic models for flood prediction and flood hazard mapping. Therefore, it is essential to calibrate and validate the channel's Manning's n roughness coefficient using simulation models. In this study, we calibrated and validated Manning's n roughness coefficient using HEC-RAS for Sarimsakli creek that is tributary of the Kizilirmak river in central Anatolia, Turkey. For calibration of Manning's n-values, six different flow regimes were considered based on mean daily flow records between 2005 and 2010. We found that calculated water surface profiles for different Manning's n are slightly lower than the measured one, that indicates a continuous underestimation of the roughness coefficient n by the model. Therefore, our results suggest that higher values of Manning's n should be considered in case of the intermittent rivers. A polynomial relationship was proposed between roughness values and Froude numbers. Finally, a linear relation was established between calculated and measured Manning's n roughness coefficient. Nevertheless, results showed that careful attention should be given to higher n-values because the differences between HEC-RAS and Manning's n becomes larger. Solution-oriented results obtained and the methodology applied to the Sarimsakli creek may serve as a practical reference for HEC-RAS modelling and flood prediction in intermittent rivers.