Scholarlink

<p>Using the Northeast China cold vortex index (NCCVI) based on the large-scale circulation pattern, this study developed a dynamic prediction model for NCCV intensity by reconstructing its nonlinear state space trajectory. The reconstructed trajectory dynamics were demonstrated to be equivalent to those of the original system generating the time series, thereby enabling the establishment of a reliable model for forecasting future system states. Three nonlinear prediction approaches were implemented: (1) a single-variable NCCVI prediction method, (2) a combined approach incorporating NCCVI with precursor signals (including sea surface temperature anomalies, snow cover, and sea ice concentration), and (3) an approach integrating NCCVI with forcing signals extracted through slow feature analysis. These methods were applied to predict monthly NCCV during warm seasons. Prediction results and sensitivity tests demonstrate that the established methods can exhibit predictive skill for monthly NCCV activity. As a valuable complement to existing dynamic and statistical methods, these nonlinear state space reconstruction techniques effectively enhance our capability to predict NCCV activities.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Prediction of the Northeast China Cold Vortex Index Based on Nonlinear State Space Reconstruction

  • 期刊论文

  • 发表日期：

    2026年6月25日

查看全文
  • Geli Wang,
  • Kaiyu Fan,
  • Qiduo Shao,
  • Zelun Cheng,
  • Zuowei Xie,
  • Cholaw Bueh,
  • Ziniu Xiao

摘要

Using the Northeast China cold vortex index (NCCVI) based on the large-scale circulation pattern, this study developed a dynamic prediction model for NCCV intensity by reconstructing its nonlinear state space trajectory. The reconstructed trajectory dynamics were demonstrated to be equivalent to those of the original system generating the time series, thereby enabling the establishment of a reliable model for forecasting future system states. Three nonlinear prediction approaches were implemented: (1) a single-variable NCCVI prediction method, (2) a combined approach incorporating NCCVI with precursor signals (including sea surface temperature anomalies, snow cover, and sea ice concentration), and (3) an approach integrating NCCVI with forcing signals extracted through slow feature analysis. These methods were applied to predict monthly NCCV during warm seasons. Prediction results and sensitivity tests demonstrate that the established methods can exhibit predictive skill for monthly NCCV activity. As a valuable complement to existing dynamic and statistical methods, these nonlinear state space reconstruction techniques effectively enhance our capability to predict NCCV activities.