利用對流尺度系集資料同化系統同化巨量雷達資料及其對豪大雨預報之影響(1/2)

專案詳細資料

Description

This proposal aims to understand the convective-scale predictability, in particular the short-live and intense afternoon thunderstorms in Taiwan. The optimal goal is to improve the associated forecast skill by improving the current convective ensemble radar data assimilation system to better represent the small-scale features and investigate how/whether these corrections improve the dynamic/thermodynamic conditions of heavy rainfall. For this goal, we plan to optimize the performance of the NCU WRF-LETKF radar assimilation system to assimilate large volume high-resolution radar data with reasonable computation efficiency under the collaboration between NCU (Taiwan) and RIKEN (Japan). To avoid overfitting and forecast degradation, the observation error correlation is included in WLRAS, recently. The new implementation is under investigation with a meiyu-front related heavy rainfall event based on four C-band radars. Preliminary results suggest that introducing correlated observation error to WRLAS leads to more small-scale corrections and improve precipitation prediction. However, further efforts are required to investigate the potential impact on the less predictable small-scale intense precipitation events over Taiwan, with the assimilation of all available (large volume) radar data in Taiwan using the high-performance computing system.
狀態已完成
有效的開始/結束日期1/04/2131/03/22

聯合國永續發展目標

聯合國會員國於 2015 年同意 17 項全球永續發展目標 (SDG),以終結貧困、保護地球並確保全體的興盛繁榮。此專案有助於以下永續發展目標:

  • SDG 13 - 氣候行動

Keywords

  • Big data analysis
  • data assimilation
  • heavy rainfall prediction
  • radar data

指紋

探索此專案觸及的研究主題。這些標籤是根據基礎獎勵/補助款而產生。共同形成了獨特的指紋。