In this graduation thesis we discuss runoff modelling from the upper part of the Gradaščica river basin. Hydrological study was performed using hydrological model HEC-HMS. This thesis presents the preparation of required inputs for hydrologic model and some basic concepts of runoff modelling, which are important for understanding of the runoff characteristics. Using freely available computer program SAGA-GIS, we defined topographical characteristics, land use, geological and pedagogical soil structure of the upper Gradaščica river basin. This data helped us to evaluate parameters of precipitation loss and transformation precipitation runoff methods. Unit hydrograph (UH) theory was used for runoff modelling. Moreover, discharge measurements at river sub-basins which were defined by the watershed coding system were also performed in the scope of the thesis.
The hydrological model was calibrated and validated using the measured rainfall and discharge data. Using manual calibration we tried to obtain the combination of parameters that gives the best correspondence between measurements and modelled discharge values. The calibration and validation process was performed using discharge data from the Dvor gauging station. Clark UH method yielded the best results and was selected as the most suitable method of unit hydrograph methodology. Snyder UH method yielded somewhat worse results. Furthermore, the SCS UH gave the worst correspondence between measured and modelled data. The CN parameter for the upper Gradaščica river basin was estimated to be from 65 to 72 for the three river sub-basins (calibrated values of the parameters), which we were defined in the process of hydrological modelling and data preparation. During validation process of the hydrological model we found that antecedent precipitation conditions have influence on the differences between measurements and modelled discharge values.
Using calibrated and validated hydrological model we also defined hydrographs with 10-, 50- and 100-years return period. For the input data we used synthetic histogram precipitation and temporal precipitation distribution was also taken into account. Modelled peak discharge values were also compared with flood frequency analysis results.