Early warning and accurate forecasting of weather and climate extremes are critical to avoid natural disasters that damage human lives, especially in megacities. Heavy rainfall can engender widespread flooding, landslides or mudslides, and urban waterlogging, causing substantial economic losses and casualties (Blöschl et al., 2017;Field et al., 2012;Gariano & Guzzetti, 2016). The warning of extreme rainfall is largely determined by the effective leading time of dynamical forecasting systems. While numerical weather prediction (NWP) is an indispensable tool for handling extreme rainfall events, forecasting on the subseasonal-to-seasonal (S2S) time scale (10-90 days) may add the value of advance warning to better preparedness and emergency measures (Merz et al., 2020;White et al., 2022).The S2S prediction project releases operational forecasts and reforecasts, with particular emphasis on high-impact weather events (Vitart et al., 2017), which provide a valuable tool for risk management and communication with