2016年1月23日受到強烈大陸冷氣團影響，台灣許多山區開始降雪，1月24日(0400 UTC)甚至在低海拔地區觀測到固態水象粒子的存在，而此亦為台灣超過50年來首次在海拔200公尺以下觀測到降雪，許多測站低溫記錄皆創下歷史新低，更導致農業、漁獲出現嚴重災情。本研究利用中央氣象局一維雷射光學式雨滴譜儀(PARSIVEL)來校驗中央大學雙偏極化雷達(C-POL)水象粒子分類(Particle identification，簡稱PID)的反演類別，並進行高空及地面觀測資料分析驗證，討論降雪時空分布特徵及水象粒子分類。

雙偏極化雷達提供高時間、高空間解析度的資料，並透過雷達參數ZHH (水平回波)、ZDR (差異反射率)、ρHV (相關係數)、φDP (差異相位差)及KDP (比差異相位差)反應不同相位粒子的特性，進而對降水粒子作出分類，研究中分別針對1月23日於鞍部測站以及1月24日於中央大學科二館頂樓觀測到固態降水粒子的時間進行分析討論。

根據雷達PPI掃描結果顯示，從1月23日1319UTC至2018UTC，0.5ﾟ低仰角掃描ZDR和RHV參數皆有一圈明顯的融解層(亮帶)，並隨時間逐漸向雷達靠近，近一步剖面發現此圈融解層隨著時間越晚以及溫度的降低越來越貼近地面，融解層的出現意味著一個冰水混相的特徵，也進階說明了此時近地面可能已有固態水象粒子的蹤跡。中央氣象局鞍部測站的光學式雨滴譜儀，在此期間亦偵測到粒子的平均落速逐漸變小，粒子的總個數增加的情形，採用Atlas and Ulbrich(1977)提出雨滴的Z-R關係式以及Locatelli and Hobbs(1974)對於不同類型的降雪粒子平均落速和粒徑的最佳曲線(best-fit curves)作為參考，發現本個案粒子平均落速曲線的下降趨勢顯示出由層狀降雨逐漸轉變為降雪的特性，也最為接近Locatelli and Hobbs(1974)水象粒子分類研究中的Densely rimed dendrites類型。而後調整前人研究中粒子粒徑-平均落速關係式，歸納出較符合台灣使用的降雪公式範圍。

最後使用雷達參數進行PID反演，以模糊邏輯為基底，針對個案建立雷達參數與水象粒子隸屬函數間的關係來判別所選區域內存在的水象粒子類型時，經調整隸屬函數後獲得的水象粒子分類結果主要是以濕雪(Wet snow)及小雨(Light rain)為主，與在鞍部測站及中央大學觀測到的水象粒子類型結果吻合。

Geographically, Taiwan is located at subtropics, snowfall rarely occurred on the low elevation except the mountains. However, January 23 to 24, 2016 northern Taiwan was affected by strong continental cold high pressure system, many low altitude areas observed solid hydrometeor particles.

In order to analyze and validate this snow event, National Central University C-POL dual-polarization radar and CWB laser-optical Particle Size Velocity (PARSIVEL) at Anbu station were employed to classify the type of solid hydrometeor particles.

Dual-Polarization radar provides both high spatial and temporal resolution. An obvious layer with high differential reflectivity ZDR and low cross-correlation coefficient ρHV at very low altitude was observed in vertical cross section as the precipitation type switched from stratiform rain to snow. Simultaneously, the terminal velocity of particles turned smaller and the numbers of aggregrate increased at Anbu parsivel station. The radar data analysis matched parsivel data in the course of melting level gradually approached ground. According to Atlas and Ulbrich (1977) present Z-R relationship and Locatelli and Hobbs (1974) found the best-fit curves between diameter and terminal velocity of particles, summarize a range of suitable snowfall formulas in Taiwan.

Through the comparison between preliminary data analysis and the documented statistical classification results of ice particles, according to Locatelli and Hobbs(1974), the particles was classified as Densely rimed dendrites type. Further, the NCAR fuzzy logic bulk-hydrometeor particle identification algorithm (PID) was applied to the radar data to get the three-dimensional distribution of particle types. Wet snow, light rain and graupel were detected as a result of PID at lowest level, these results conformed with the in situ observation at NCU and Anbu station.

參考文獻

劉妍利，2007：梅雨降水系統的雙偏極化雷達資料分析與WRF模式模擬研究,國立中央大學大氣物理所碩士論文。1-26

林育邦，2015：應用雙偏極化氣象雷達反演過冷水之個案分析。6-7

楊錫豐，2013：颱風雨帶水象粒子分布特徵研究。13-16
Crystal J., Lawrence D., Walter A., Petersen, Mariana, William. ,2011：Development and Testing of Operational Dual-polarimetric Radar Based Lightning Initiation Forecast Techniques. NOAA Tech. Memo. 7p

Edward A., Brandes, Kyoko Ikeda, Gregory Thompsonmichael Schonhuber,2008：Notes and Correspondence Aggregate Terminal Velocity/Temperature Relations. J. Atmos. tec. 2729-2736

Elizabeth J., Steven A., Brenda ,2014：A Dual-Polarization Radar Hydrometeor Classification Algorithm for Winter Precipitation. J. Atmos. tec.,1457-1481

Keenan,2002: Hydrometeor classification with a C-band polarimetric radar. Aust. Met. Mag,52, 23-31

Locatelli J.D., Hobbs P.V.,1974：Fall Speeds and Masses of Solid Precipitation Particles. Journal Geophysical .185-2197

Marzano, Scaranari D., Celano M., Alberoni P., Vulpiani1 G., Montopoli M., 2005: Hydrometeor Classification From Dual-Polarized Weather Radar:Extending Fuzzy Logic From S-band To C-band Data. Advances in Geosciences, 7, 109–114, 2006

Takeharu,K., Hiroshi,U.Tadayasu,O.,Mariko,O.,2015：A Hydrometeor Classification Method for X-Band Polarimetric Radar：Construction and Validation Focusing on Solid Hydrometeors under Moist Environments. J. Atmos. ocean,2052-2074

Vivekanandan, J., Zrnic, Ellis, Oye R.., Ryzhkov , Straka, 1999：Cloud Microphysics Retrieval Using S-Band Dual- Polarization Radar Measurements. Bull.Amer. Meteor.Soc., 80, 381-388.