Compound events, commonly defined as the “combination of multiple drivers and/or hazards that contributes to societal or environmental risk”, often result in amplified impacts compared to individual hazards. The outputs of climate models are used to understand the evolution of compound events in terms of frequency in a climate change context. Climate models are known to be statistically biased, hence 2 multivariate bias correction methods (dOTC and R2D2) and one univariate bias correction method (CDF-t) are compared to uncorrected CMIP-6 data over return periods of compound events, and extremal dependence structure. This study focuses on two compound floodings triggered by accumulated precipitation and modeled either with a Multivariate Generalized Pareto (MGP) approach or a copula-based approach. First, it is shown that the representation of the two compound floodings in climate models are biased, and that bias correction methods improve this representation. Then, bias correction methods, calibrated on ERA5 data over the period 1992-2021, are applied to the future in order to assess the evolution of the two compound floodings in the future. Extension to Europe are finally discussed.