Hydrological Modeling in Northern Tunisia with Regional Climate Model Outputs: Performance Evaluation and Bias-Correction in Present Climate Conditions

摘要: This work aims to evaluate the performance of a hydrological balance model in watershed located northern Tunisia (wadi Sejnane, 378 km2) present climate conditions using input variables provided by four regional models. A modified version (MBBH) lumped and single layer surface BBH (Bucket with Bottom Hole model, which pedo-transfer parameters estimated physiographic characteristics are introduced) is adopted simulate water components. Only two representing respectively retention capacity soil vegetation resistance evapotranspiration calibrated rainfall-runoff data. The evaluation criterions for MBBH calibration are: relative bias, mean square error ratio actual potential evapotranspiration. Daily air temperature, rainfall runoff observations available from 1960 1984. period 1960–1971 selected while 1972–1984 chosen validation. Air temperature precipitation series models (DMI, ARP, SMH ICT) European program ENSEMBLES, forced global (GCM): ECHAM ARPEGE. outputs (precipitation temperature) compared terms statistical distribution. analysis was performed at seasonal scale precipitation. We found out that RCM must be corrected before being introduced as inputs. Thus, non-parametric quantile-quantile bias correction method together dry day employed. Finally, simulated generated most acceptable comparison observed data, reproduce temporal variability monthly runoff. accurate occurrence days but still underestimates them. From distribution point view, data into were not able extreme values during validation (larger than 80 mm/month). may due weakness reproducing moderate high levels even after correction. approach considered way use (RCM) studying impacts.