A translation wave model: Güneycedere case study

Abstract

Abstract. This study proposed a flood routing model which was derived from Saint-Venant (SV) equations. It can be called translation wave model (TWM). In this model, bed slope term and friction slope term were ignored in the momentum equation of SV equations. This means, the difference between bed slope and friction slope are relatively small compared to other terms in the SV equations. This approach is similiar to the one in kinematic wave model (KWM), but in KWM inertia and pressure terms are neglected. In this study, governing equations for the proposed model were derived and solved numerically by using an explicit scheme. Then, validation of the proposed model was obtained through real flood data that belong to an actual creek reach in Isparta Province, Turkiye. The creek reach was between two stream gauging stations and the inflow and outflow hydrographs of a real flood event were available. Also, KWM was implemented for this creek reach using this real flood event. Thus two simulated outflow hydrographs; one that belongs to KWM and another that belongs to TWM were created. Then the two simulated outflow hydrographs were compared by differences in peak discharge, time to peak flow and hydrograph volume. Since KWM fails to predict attenuation and dispersion in outflow hydrographs, relative error of peak flow in KWM is calculated bigger than in TWM (2,19 % > -0,27 %). Relative error of time to peak flow in TWM is calculated as 0,00 % while it is calculated -2,50 % in KWM and the two models failed to provide volume conservation. Also, TWM and KWM were evaluated by the statistical parameters; Root Mean Square Error (RMSE), Mean Absolute Error (MAE) and Nash-Sutcliffe Efficiency (NSE). The results were in acceptable range but KWM gave better results since the creek reach had a steeper slope than average (S0 ≥ 0.005). Finally, for comparison, an inflow hydrograph from literature was routed with KWM and TWM in a rectangular channel.