Despite its pronounced global impacts, tropical Pacific decadal variability (TPDV) is poorly predicted by current climate models due to model deficiencies and a limited understanding of its underlying mechanisms. Using observational data and a hierarchy of model simulations including decadal hindcasts, we find that decadal isopycnal depth variability driven by oceanic Rossby waves in the tropical Pacific provides the most important source of predictability for TPDV. The predictability arising from initial isopycnal depth conditions is further amplified by tropical ocean-atmosphere coupling and variations in the strength of subtropical cells in the Pacific throughout the decadal forecasts. Regional initialization experiments that effectively isolate the impact of different ocean basins on TPDV predictability highlight the essential role of the tropical Pacific. This study enhances our understanding of the mechanisms governing TPDV predictability, offering crucial insights for improving the accuracy of decadal predictions.

尽管热带太平洋年代际变率 （TPDV） 具有显著的全球影响，但由于模型缺陷和对其潜在机制的理解有限，当前的气候模型对热带太平洋年代际变率 （TPDV） 的预测效果不佳。使用观测数据和包括年代际后报在内的模型模拟层次结构，我们发现热带太平洋海洋罗斯比波驱动的年代际等密度深度变化为 TPDV 提供了最重要的可预测性来源。在整个年代际预报中，热带海洋-大气耦合和太平洋副热带单体强度的变化进一步放大了初始等密度深度条件产生的可预测性。有效隔离不同海洋盆地对 TPDV 可预测性影响的区域初始化实验突出了热带太平洋的重要作用。这项研究增强了我们对控制 TPDV 可预测性机制的理解，为提高年代际预测的准确性提供了重要的见解。