The “Predictability Barrier” Phenomenon of Winter Extreme Cold Events in Central and Eastern China and Mechanisms of Error Amplification

Abstract Previous studies have primarily focused on evaluating the forecast skill of extreme cold events in central and eastern China as a whole, with limited attention to their different stages. This study identifies a distinct “predictability barrier” phenomenon in the ensemble forecasts, characterized by rapid growth of ensemble mean forecast error in 2m‐temperature during the intensification stage of the events. In contrast, the forecast error tends to decrease during the decay stage. Consequently, the decay stage is more accurately forecasted than the intensification stage at the same lead time. Mechanism analyses indicate that error amplification is primarily driven by the interaction between the horizontal wind forecast error and the background horizontal temperature gradient of the event, which is dominantly governed by event intensification. Error reduction during the decay stage is primarily dominated by the conversion of available potential energy error into kinetic energy error.