:::
104/10/06 14:00 中央大學大氣科學學系楊舒芝副教授 演講
演講公告
張貼人：網站管理員 ╱ 公告日期：2015-09-30演 講 公 告
主講人：楊舒芝 副教授
地點：中央大學科學一館S-325教室
摘要：
講題：利用NCU Regional Assimilation System (NCU-REAS)改善台灣區域劇烈天氣預報
Improving severe weather prediction in Taiwan with the NCU regional ensemble assimilation system
主講人：楊舒芝 副教授
中央大學大氣科學學系
時間：10月06日（星期二）下午2點
地點：中央大學科學一館S-325教室
摘要：
本研究透過個案分析利用NCU區域系集同化系統(NCU-REAS)探討系集卡爾曼率波器在針對如梅雨或颱風劇烈天氣型態時所可能面對的難題及可能解決方案。
為掌握梅雨期間，西南氣流水氣傳輸及台灣地形交互作用下所引起的強對流降水，我們利用雙局地化法在高解析度模式網格點下進行多重尺度修正，並以2008年SoWMEX IOP8 6月16日豪大雨個案為例進行討論。結果顯示，雙局地化法能保留在南中國海面上縱貫尺度水氣傳輸修正，且能保留台灣西南沿岸及平地局地風場特性，因此有助於改善降雨預報的位置及強度。
針對颱風預報，我們在NCU-REAS的架構下建立颱風同化策略：包含系集渦漩中心一致化及渦漩置換，目的便是凸顯颱風真實結構的不確定性，並以此改善與颱風相關的背景場誤差結構。以2010凡那比颱風為例，此颱風同化策略有助於改善強度預報，但對路徑預報影響不大。
The NCU regional ensemble assimilation system (NCU-REAS) is established by coupling the Local Ensemble Transform Kalman Filter with the Weather Research and forecasting (WRF) model. This system has been applied to study the issues related to severe weather prediction in Taiwan. In this talk, I will present recent improvements with NCU-REAS, including the multi-scale multi-resolution assimilation for heavy precipitation prediction and the storm-centered framework for typhoon prediction.
The multi-scale multi-reolution framework is developed based on the dual-localisation method and the WRF nested domains. With the heavy precipitation episode on 16th June 2008, our results show that the dual-localization method can improve the intensity and location of the heavy rainfall prediction by preserving the synoptic-scale moisture transport by the southwesterly monsoonal flow over South China Sea and resolving the strong convective-scale motion in high-resolution grids.
For typhoon prediction, a new tropical cyclone assimilation framework is constructed based on vortex replacement and a storm-centered method. With the 2010 Typhoon Fanapi, our results show that such assimilation framework can better represent the flow-depedent background errors associated with TC dynamics and particularly bring positive impact on intensity prediction.
最後修改時間：2015-09-11 PM 3:30
