The uncertainty of flood frequency analyses in hydrodynamic model simulations

Abstract. Assessing the risk of a historical-level flood is essential for regional protection and resilience establishment. However, due to limited spatiotemporal coverage observations, impact assessment relies on model simulations thus subject uncertainties from cascade physical processes. This study assesses hazard map with different combinations runoff inputs, variables frequency analysis fitting distributions based estimations by CaMa-Flood global hydrodynamic model. Our results show that deviation in inputs most influential source estimated flooded water depth inundation area, contributing more than 80 % total investigated this study. Global maps floods 1-in-100 year return periods large uncertainty values but small ratios river channels lakes, while opposite are found dry zones mountainous regions. result increasing variation at tails among various distributions. In addition, between selected increases regular period rarer floods, both points area over The also lead estimating population economy exposure floods. total, accounts 9.1 [8.1 %–10.3 %] land flood, leading 13.4 [12.1 %–15 13.1 [11.8 %–14.7 economic globe. vary continent Africa have largest uncertainty, probably observations constrain simulations, indicating necessity improve performance hydrological models especially data-limited