Robust decadal predictability of cold surge frequency in Taiwan and East Asia through teleconnection of North Atlantic Oscillation

This study investigates the decadal predictability of cold surge frequency (CSF) in East Asia, including Korea, Japan, and Taiwan, through the lens of the North Atlantic Oscillation (NAO) index. The findings suggest that extreme events such as cold surges can be predicted on decadal timescales when the teleconnection mechanism is robustly established. The study revisits and consolidates the dynamical mechanisms underlying wave propagation and the teleconnection between the NAO and the East Asian trough, highlighting their role in creating a winter environment conducive to cold surges in Taiwan and East Asia. The study demonstrates the skill of climate models in capturing the NAO's decadal variability, and develops a statistical-dynamical hybrid approach. This method integrates decadal prediction datasets with a statistical model to enhance the prediction of extreme cold surge occurrences on a multi-annual timescale. The results of the study underscore the scientific significance of merging climate dynamical mechanisms with decadal prediction systems for extreme events, and introduce a hybrid framework that combines numerical decadal climate predictions with statistical regression models. This addresses the challenges posed by biases in climate prediction models and advances the capability to predict regional extreme events such as cold surges.