The traditional method of calibrating a computational river model is to adjust the model parameters in a systematic manner until a reasonable agreement between observations and computed values is obtained, This process is subjective and can be time-consuming, hence the development of an automatic calibration procedure has been of particular interest in recent years, Automatic calibration procedures require a means of comparison agreement between observations and the model simulations called the ＇objective function＇. This paper presents an objective function a involving a weighting function that is tuned to matching key features of the flow hydrograph, Eleven objective functions were developed and the best one selected from assessing their performance on an artificial model that produced a hydrograph with two peaks, The best objective function was subsequently used to calibrate the flow resistance automatically in a hydrodynamic one-dimensional river model based on an explicit finite difference numerical scheme.