Quantifying the uncertainty associated with future floods simulations from regional climate models

This study is motivated by showing how future floods predictions in arid areas respond to uncertainty in climatic parameters—a question, if explored, that bridges a gap in water resources management plans. To address this question, we projected the changes, over Qatar, in four climatic parameters (surface air temperature, precipitation, wind speed, and potential evapotranspiration) from eight Regional Climate Models (RCMs) under two Representative Concentration Pathways (RCP4.5 and RCP8.5) during the middle of the 21^st century (2031–2050). Using topographic and groundwater data, a physically-based water balance model was built to simulate future floods under all these scenarios. Findings show high uncertainty in climatic parameters. Relative to the historical period, values varied under RCP4.5 (RCP8.5) from +1.8 to +3.4 (+3.8 to +5.6)^°C for average temperature, -48% to +15% (-60% to +6%) for annual precipitation, -0.23 to +0.1 (-0.27 to +0.04) m/hour for wind speed, and from -5.7 to +12.8 (+4.3 to +17) mm for annual potential evapotranspiration. Uncertainty in climatic parameters caused significant uncertainty in future floods estimations. During the middle of the 21^st century, floods simulations varied from -67% to +64% with an average value of -20% under RCP4.5, and from -81% to +8% with an average value of -36% under RCP8.5. The greatest uncertainty resulted from the driving models, whereas the choice of emission scenario had a secondary impact. Since floods studies are critical to save lives and assets, the study’s findings emphasize the importance of both considering the uncertainty associated with climatic parameters and the regional climatic information chosen.