本研究透過觀測及系集模擬的結果，探討2022年6月7日梅雨鋒面沿海強降雨個案之特徵及機制，進而辨認較易發生強降雨之環境，以及深入討論導致此降雨事件的主要因素及其相關性。利用多都卜勒風場合成系統(Wind Synthesis System using Doppler Measurements, WISSDOM)得到的高解析度三維風場，分析動力場的時間演變。系集敏感度分析(Ensemble Sensitivity Analysis, ESA)用於說明熱動力的變數和降雨之間在時空上的關聯性，而k-Means群集分析能夠有效的對於系集成員的熱動力特徵和降雨的相似性做分類。反演風場的結果顯示受到臺灣地形的影響，產生之地形噴流及鋒前強西南風的偏轉導致的底層輻合，提供降雨的動力條件。在系集模擬的分析中，由ESA的結果得到風場及水氣對於沿海的降雨較為重要，而群集分析則將系集成員分類並凸顯具有適合降雨環境的群集。由觀測和模擬的結果顯示大尺度的熱動力特徵和中小尺度的動力機制，皆為本研究之沿海降雨的形成之重要因素。;This study investigated the features and mechanisms of coastal heavy rainfall associated with the Mei-Yu front using numerical simulations on June 7, 2022, and identified whether favorable environments corresponded with similar precipitation in ensemble simulations, exploring the relationship between the primary factors that result in precipitation. Ensemble Sensitivity Analysis (ESA) illustrated the relationship between dynamic, thermodynamic fields and precipitation, and k-means Clustering classified the ensemble members with similarities on precipitation or model variables. The synthetic wind showed a southwesterly wind along the terrain during the rainfall period. The strong coastal low-level convergence between westerly and southerly winds was caused by the terrain. In ensemble simulations, the large-scale environment controlled by wind and water vapor, according to ESA, which were important in producing heavy rainfall near the coast. The clustering results separated members with better rainfall performance and favorable rainfall conditions effectively. Both the large scale frontal system and mesoscale dynamic processes contributed to the formation of coastal heavy rainfall.
