Drivers and predictability of summer marine heatwaves in the Northwest Pacific

Marine heatwaves (MHWs) in the Northwest Pacific (NWP) exert significant ecological and climatic impacts, yet their drivers and predictability are not fully understood. Based on the multivariate empirical orthogonal function (MV-EOF) method, this study identifies two dominant modes of summer NWP MHWs. The first mode, characterized by widespread warming across the low-latitude NWP, occurs during the summer following an El Niño event. This mode is strongly associated with the Pacific–Japan teleconnection pattern and sea surface temperature (SST) anomaly gradient between the North Indian Ocean and tropical western Pacific. The second mode exhibits a northeast-to-southwest tripole structure, representing the summer phase during El Niño development. This tripole pattern is possibly influenced by the North Pacific Oscillation, highlighting an extratropical–tropical teleconnection that propagates the effects of positive SST anomalies. Using physics-based empirical prediction models validated by the leave-one-out cross-validation approach, a notable degree of predictability is found for these MHW modes. The temporal correlation coefficient scores and Root Mean Square Errors between observed and predicted principal components (PC1 and PC2) reach 0.65 and 0.77 for PC1 and 0.55 and 0.84 for PC2, respectively, over the period 1982–2022. Both models effectively capture the peak intensity and spatial distribution of MHWs, despite minor discrepancies. These findings might advance our understanding of MHW dynamics in the NWP and provide a foundation for developing early warning systems to mitigate their adverse effects on marine ecosystems and coastal communities.