Quantifying devastating weather events under climate change using multivariate bias correction

Some of the most devastating climatic events of recent decades involve the combination of several climatic variables on fairly large scales and over a period of time. Sometimes combinations of elementary climatic events which are not necessarily devastating when isolated can have devastating effects when they occur simultaneously or successively within a short period.

These extreme climatic events, called compound events (EC, Zscheischler et al., 2020) correspond to particular combinations of climatic variables presenting patterns of spatio-temporal dependencies and between the variables. These events can be extracted from reanalysis data (from the past period to the current period) as well as climate models.

Due to discretization effects, low spatial resolution or implemented simplifications, climate models exhibit various biases in their marginal, spatial and inter-variable properties, which necessarily affect the representativeness of compound events. These biases must be corrected using bias correction methods. The objective of this thesis project is to establish a statistical relationship between high-impact catastrophic compound events and the large-scale structures that are their triggers, first on reanalysis data in the near past, and then to project this relationship in the future to assess these catastrophic events in terms of frequency, magnitude and impact. Considering the shortcomings of current bias corrections, several multivariate bias correction methods will be tested and compared in the near past (as well as compared to univariate correction, as a benchmark) and the best correction method will be selected for projection. .