為掌握劇烈天氣如梅雨，颱風系統預報，雲解析模式發展，高時空解析度觀測及相對應高解析度模式同化系統及預報為當代數值天氣預報重要發展趨勢。以台灣而言，劇烈天氣豪雨及強風預報的問題更為複雜。尤其在台灣北部系統性劇烈天氣在接近台灣北部之移動速度，環境場影響，局地環流特性，模式物理過程等複雜多重尺度交互作用使可預報度更低。本計畫希望透過台灣區域大氣水文與海洋觀測計畫(TAHOPE，詳細說明請見整合計畫)及台美觀測實驗計畫(PRECIP2020)之密集觀測及新一代衛星觀測資料等，探討多重尺度交互作用下影響台灣北部之劇烈天氣系統風力及降雨等強度相關之預報敏感度，及觀測可扮演角色。此外，考慮劇烈天氣系統之濕度場準確度對劇烈天氣系統之發展有重要的影響，本研究亦將針對多尺度交互作用下濕度場不確定性進行探討，並評估水氣相關觀測之重要性。主要將透過高解析多尺度資料同化系統，預報敏感度估計工具探討台灣北部劇烈天氣系統強度之可預報度，並期應用於實際對流資料同化系統發展並進一步改進強度預報。在實驗準備階段將先以觀測系統實驗及觀測系統模擬實驗針對豪雨及強風個案進行同化與預報實驗。 ;Heavy rain and strong wind prediction in Taiwan is very challenging. Moreover, the predictability of severe weather in northern Taiwan is even more complicated due to the movement/development of the severe weather system, interaction with the environment or local circulation, microphysics and complex terrain. The convective-scale data assimilation and prediction actually involves forecast errors multi-scales. The proposal aims to study the predictability of the severe weather in northern Taiwan based on the observations collected during TAHOPE and PRECIP2020 and through the high-resolution, multi-scale data assimilation system. Particularly, we will focus on the assimilation of moisture observations and study the impact of moisture errors on the development of severe weather systems. As a preparation for TAHOPE, issues related to intensity prediction will be investigated based on the observation system experiments and observation system simulation experiments.
