Seasonal Prediction of Extreme High-Temperature Days in Southwestern China Based on the Physical Precursors

Extreme high temperatures frequently occur in southwestern China, significantly impacting the local ecological system and economic development. However, accurate prediction of extreme high-temperature days (EHDs) in this region is still an unresolved challenge. Based on the spatiotemporal characteristics of EHDs over China, a domain-averaged EHD index over southwestern China (SWC-EHDs) during April−May is defined. The simultaneous dynamic and thermodynamic fields associated with the increased SWC-EHDs are a local upper-level anticyclonic (high-pressure) anomaly and wavy geopotential height anomaly patterns over Eurasia. In tracing the origins of the lower boundary anomalies, two physically meaningful precursors are detected for SWC-EHDs. They are the tripolar SST change tendency from December-January to February−March in the northern Atlantic and the February−March mean snow depth in central Asia. Using these two selected predictors, a physics-based empirical model prediction was applied to the training period of 1961–2005 to obtain a skillful prediction of the EHDs index, attaining a correlation coefficient of 0.76 in the independent prediction period (2006–19), suggesting that 58% of the total SWC-EHDs variability is predictable. This study provides an estimate for the lower bound of the seasonal predictability of EHDs as well as for the hydrological drought over southwestern China.