Hydrological modeling based on ensemble prediction system in the mountainous Euphrates basin

Subjects such as increasing world population, global warming, flood and drought, increase the interest in water resource management and hydrology science. Therefore, in the mountainous Euphrates Basin, hosting large reservoir dams, runoff modeling and forecasting during spring and early summer is important in terms of energy and water resources management. In this study, the Upper Euphrates (Karasu) Basin (10,275 km2) located at the headwater of Euphrates River, one of Turkey's most important river, is selected as an application area. The aim of the study is to make forward-oriented, medium-range flow forecasts using probability based forecast system. Conceptual hydrological model HBV, which has a common usage in the literature, is chosen to predict streamflows. Ensemble Prediction System (EPS) including 51 different weather forecasts produced by ECMWF is used as probability based weather forecast data. Melting period during March-June of 2008-2012 is chosen as the study period. First of all, performance evaluation analysis is done to determine the consistency of archived EPS forecast data with ground observations and bias correction is performed to resolve the systematic errors in Ensemble Prediction System. Renewed performance analysis showed an increasing consistency between forecast data and ground observations. Then, using raw and corrected EPS data in HBV, runoff forecasts are done for 1-9 days lead time. As a result, ensemble and simulated runoff comparisons are presented and discussed. This thesis sets a pioneer application for the probabilistic runoff forecats based on EPS in Turkey and is supposed to lead operational hydrological applications and water resources management. Keywords: Ensemble Prediction System; Euphrates Basin; HBV; Bias Correction