Multiscale Drivers and Subseasonal Forecast Skill of Persistent Heavy Rainfall Events during the Record-Breaking 2025 North China Rainy Season

The 2025 North China rainy season was record-breaking, with accumulated rainfall 132% above climatology and a duration 28 days longer than normal. Rather than reflecting season-long wet conditions, this anomaly was produced by three persistent heavy rainfall (PHR) events, denoted as PHR1, PHR2, and PHR3, during the onset, mature, and withdrawal stages, respectively. Using observations, reanalysis, and ECMWF subseasonal-to-seasonal (S2S) ensemble forecasts, this study examines how distinct timescale-dependent processes shaped these events and their forecast skill. PHR1 was associated with the interaction between a 10–30-day tropical intraseasonal oscillation and a midlatitude wave train, which jointly established a southwesterly moisture corridor toward North China. PHR2 featured multiscale interactions, with 10–40-day intraseasonal vertical moisture advection making the largest contribution to rainfall development. PHR3 was linked to northward propagation of the 30–90-day boreal summer intraseasonal oscillation (BSISO), which helped sustain anomalous moisture transport and rainy-season-favorable circulation in late August. ECMWF forecasts show clear event-to-event differences in skill. PHR1 was poorly predicted because of limited skill in capturing the 10–30-day tropical–extratropical intraseasonalos-cillation (ISO) configuration. PHR2 exhibited large ensemble spread associated with uncertainty in intraseasonal vertical motion. PHR3 showed the highest skill, owing to better prediction of low-frequency BSISO. Despite these differences, realistic prediction of tropical intraseasonal evolution emerged as a key condition for useful extended-range forecasts. These results reveal how a record-breaking rainy season can be built by scale-dependent heavy rainfall events and highlight tropical intraseasonal variability as a key source of subseasonal forecast skill over North China.