Extreme future precipitation in Bologna: an exploration based on different weighted multi-model ensemble methods

Assessing the statistical behavior of future extreme precipitation is a topical issue for the mitigation of pluvial and flood risk. There is increasing evidence that extreme short-duration precipitation is intensifying, but the quantification of such increase is still a challenging issue. Using one of the longest available daily precipitation series—continuously recorded in Bologna since 1 January 1813—we applied five extreme precipitation indices (Rx1day, R99p, R10mm, R20mm, and R99d) to evaluate the ability of 22 bias-corrected CMIP6 climate models in reproducing historical precipitation statistics. On this basis, we compared a dynamic weighted multi-model ensemble (DW-MME) based on multi-objective Pareto optimization with an equal-weighted multi-model ensemble (EW-MME) and individual models. We further assessed the performance of the DW-MME in projecting XXIst century changes under different emission scenarios. The results show that the DW-MME provides a substantially more robust and credible representation of extreme precipitation than both the EW-MME and single-model simulations. Under high emission scenario, future extremes exhibit a clear more extreme response, with the precipitation distribution shifting toward stronger and more extreme events, revealing a pronounced dependence on climate forcing.