Advanced risk-based event attribution for heavy regional rainfall events

Risk-based event attribution (EA) probabilistically estimates changes in the likelihood of specific weather events due to global warming and is useful for climate adaptation. Unlike temperature extremes, heavy rainfall often depends on mesoscale systems and regional geography that current global circulation models (GCMs) cannot resolve. Using large-ensemble GCM simulations and high-resolution 20-km non-hydrostatic regional climate model (RCM) downscaling driven by boundary conditions from all available GCM ensemble members, we show anthropogenic warming increased the risk of two record-breaking regional heavy rainfall events in 2017 and 2018 over western Japan. Events are examined using RCM-simulated rainfall statistics and GCM-simulated background large-scale circulation. In 2017, steep terrain and a synoptic pressure pattern reduced dynamical uncertainty; in 2018, a synoptic pressure dipole favored moisture transport, amplifying risk under warming. Successful risk-based EA for regional extreme rainfall depends on how much dynamical uncertainty is reduced by background conditioning.

Yukiko Imada, Hiroaki Kawase, Masahiro Watanabe, Miki Arai, Hideo Shiogama, Izuru Takayabu