Investigating performance of the conceptual models in river hydrologic simulation

Rainfall-runoff hydrological models are important tools in water resources projects. Generally, performance of this group of models is dependent on the proper selection of parameters. Accordingly, several methods have been developed to estimate hydrological parameters. The present study aimed to compare the performance of conceptual hydrologic models such as TANK, SIMHYD and AWBM which benefit from the indirect model parameters estimation approach in discharge simulation of Babolroud watershed, Mazandaran province, Iran. The automatic calibration process of these models was designed using genetic evolutionary search algorithm and objective functions (NSE and RMSE) as error thresholds determinants. Hence, meteorological and hydrological data consist of temperature, evapotranspiration, precipitation and discharge (in daily scale) were gathered from authorities. Input data was also divided into warm-up, train and test steps after preliminary validation and recovery. Based on the results, NSE metric introduced TANK model as the best simulator respectively for train and test step (0.59 to 0.72). Depends on RMSE metric, SIMHYD (0.83) and TANK (0.15) models were introduced as the best simulator respectively for train and test step either. According to the catchment flow signatures, general simulation of low-flow (excluding the Model TANK), mean-flow and high-flow were conducted with acceptable agreement. While simulation of the flow duration curve slope which represents an intensity of changes (excluding TANK model in train step), did not provide acceptable results. Given the weaknesses and strengths of the proposed models, they can be used as an acceptable simulator in water resources management especially in terms of ungauged basins, after preliminary verification in different climatic conditions.