Constraining Subseasonal Forecast Skill for Spring 2024 Extreme Rainfall in South China by Extratropical Atmospheric Initial Conditions

Subseasonal predictability of extreme rainfall events over East Asia is determined by both atmospheric initial conditions and external boundary forcing. However, their feedback processes remain poorly understood during boreal spring. In April–May 2024, South China experienced record-breaking extreme rainfall in the decaying phase of the El Niño event. This extreme event is driven by significant crossing-time scale interactions, which offers a unique opportunity to evaluate their subseasonal predictability in dynamic models. Our results reveal that the western Pacific subtropical high (WPSH) enhanced by El Niño decaying represents the response to slowly varying oceanic forcing. In contrast, quasi-biweekly Rossby wave trains in the extratropics and the eastward-propagating Tibetan Plateau vortex reflect the atmospheric initial conditions. These extratropical atmospheric disturbances reduce the anomalous WPSH via positive advection of anomalous vorticity, thus offsetting the influences of oceanic forcing. Although the anomalous WPSH and above-normal rainfall pattern exhibit high predictability on the monthly timescale, the skillful forecasting of the three extreme rainfall events is less than 1 week. These findings suggest that the extratropical atmospheric disturbances, as initial conditions, can undermine the predictability stemming from tropical air-sea interactions. The forecast skill of the extreme rainfall process in spring is limited by the memory of initial atmospheric states and the intrinsic constraints of weather predictions.