Climate Change assessment to quantify current and future hydrological behavior of a small Mediterranean catchment

As a hotspot for climatic change, the Mediterranean basin is already struggling with the deficiency of available water resources which is expected to front with an important increase in temperature and a decrease in precipitation through the 21st century, according to the fifth report of the IPCC (Intergovernmental Panel on Climate Change). Effective assessments of climate change impact on flow regime require reliable climate model simulations and a robust hydrological impact model validated at catchment scale for variable current climate conditions. We investigate, in this work, the plausible climate change impacts on the hydrological behavior at the Raghay catchment, one of the main catchments in the high Medjerda valley in the Northern of Tunisia. We implemented a modified version of a spatially distributed conceptual hydrological model available in the ATHYS (ATelier HYdrologique Spatialisé) platform for both current and future climate conditions, taking into account the antecedent soil moisture with the integration of the Potential Evapotranspiration (PE). First, we calibrated the model using historical climate conditions and based on five accuracy criteria. An 18-year period (1990_2008) of daily rainfall, PE and discharge was used to catch a large range of past hydro-climatic conditions. Afterwards, we developed a Multi-Model Mean (MME) approach for climate change and evaluated its influence on several hydrological indicators related to water balance, magnitude and frequency of the flow for the Raghay catchment using an ensemble of five Global Circulation Models (GCMs) dynamically downscaled with a Regional Climate Model (RCM) derived from the high resolution CORDEX projections framework. In the purpose to further minimize uncertainties related to the climate projections, we performed a comparison between two bias correction methods, such as the quantile-quantile mapping (QQ) and the linear scaling (LS), of the historical precipitation projections in order to