Developing a framework for attribution analysis of urban pluvial flooding to human-induced climate impacts

Abstract

This study proposes a framework for the attribution analysis of urban flooding to historical anthropogenic climate impacts using real (historical) and counterfactual (natural) simulations from the Coupled Model Intercomparison Project Phase 5 (CMIP5) and Phase 6 (CMIP6) general circulation models (GCMs). A cluster analysis is implemented to select independent GCMs and to decrease the high computational time of the urban flood modeling. A two-step spatiotemporal statistical downscaling based on the generalized extreme value (GEV) distribution parameters is performed to meet the requirement of a fine-scale precipitation resolution for the urban flood simulations. Using the downscaled distribution parameters, the intensity–duration-frequency (IDF) curves and composite design storms for four return periods of 2, 10, 20 and 50 years are created to be used as the inputs of an InfoWorks ICM hydrodynamic model for a case study in Antwerp city, Belgium. The fraction of attributable risk (FAR) framework and the causal counterfactual probability theory are used to quantify anthropogenic climate influences on urban pluvial flooding. The results show that the GEV parameters of location and scale nicely follow the power relation, providing an opportunity for the temporal downscaling of precipitation extremes. A disagreement is found between the GCMs on the contribution of anthropogenic climate factors on pluvial flooding, some illustrating a higher probability of flooding due to anthropogenic climate impacts, and some a lower one. It can therefore be concluded that the urban flooding in Antwerp is not attributable to anthropogenic climate impacts.