雨滴譜儀觀測可以獲得雨滴粒徑分布(DSD)特徵,配合雙偏極化雷達高時空解析度的觀測,使用相關公式即可反演大範圍的DSD參數。以往使用長期統計之DSD特徵作為反演的依據,但DSD的特徵會因時間、空間、降雨類型的不同而改變,故一時一地觀測的DSD特性並不能一體適用在其他地區或時間。了解雨滴粒徑分佈的變化有助於改善反演DSD參數的成效。 本研究使用2016年9月1日到9月10日,雙北都會區夏季暴雨觀測預報實驗(TASSE)期間,五股的移動式X波段雙偏極化都卜勒雷達(TEAM-R);新店、翡翠水庫撞擊式雨滴譜儀(JWD)的觀測資料進行研究。Gamma形式的雨滴粒徑分布可由形狀參數μ、斜率參數Λ及截距參數N_0描述。偏極化參數可由Gamma參數經散射模擬求出,過程中需使用形狀參數及斜率參數的約束關係。 本研究中DSD反演所需的μ-Λ關係式使用本次個案9/8~9/10翡翠水庫JWD 統計所得的關係式及利用Brandes在2003年針對對流系統統計的μ-Λ關係式做比較。使用以上兩種關係式以TEAM-R觀測反演對流胞DSD參數的三維結構。另外再加入以JWD計算之DSD參數、JWD使用散射模擬之雷達參數直接擬合的關係式做討論,代入雷達觀測資料可直接求出DSD參數。本研究此用以上三種方法討論反演DSD參數的成效,評估後兩種使用μ-Λ關係式所得的反演結果較佳。 本研究探討2016 TASSE實驗中九月九日觀測到的對流胞,反演後得到對流胞的DSD三維結構能有效的提供對流發展時的雲微物理資訊。在對流初生時觀察到較窄的雨滴粒徑分布及較小的雨滴、較低的雨滴濃度;對流成熟時觀察到較寬的粒徑分布及較大的雨滴,雨滴濃度則比對流初生期高數百至數千倍;在層狀降雨時觀察到偏窄的雨滴粒徑分布、與較低的雨滴濃度,雨滴大小則介於對流初生期及對流成熟期之間。兩方法得到雲物理特徵十分一致,證實雨滴譜儀與偏極化雷達能有效的反演DSD的結構相當符合雲物理機制。 ;Polarimetric radar system provide high temporal and high-spatial resolution data and disdrometers can show the characteristic of Drop Size Distribution (DSD). Base on the information of disdrometer data, dual Polarimetric Radar data can be used to retrieve the parameters of DSD. The DSD varies from case to case or even in the different periods of a case. Therefore, understanding the variety of DSD is helpful to improve the accuracy of DSD parameters retrieved by radar.
The data of Taiwan Experimental Atmospheric Mobile Radar (TEAM-R) and Impact Disdrometer (JWD) observations collected in Taipei Summer Storm Experiment (TASSE) during early September of 2016 was used. From the disdrometer data, the three Gamma distribution parameters including the intercept N_0, the shape (μ) and rate (Λ) parameters can be calculated through moment method. Through the scattering calculation the relations between polarimetric variables and gamma distribution parameters can be derived. The constrain relation between the shape (μ) and rate (Λ) parameters is necessary to retrieve the three Gamma distribution parameters from polarmetric variables Z_DR and K_dp. In this study, two constrain relations were applied: first one is the well known relation proposed by Brandes et al, , the second one is from the statistics during the three days disdrometer observation of thunderstorms in TASSE. The retrieved three dimensional distribution of DSD are compared between these two retrieval applications. A third approach is directly fitting the DSD Gamma parameters with the polarimetric variables to retrieve three Gamma parameters from fitting functions. Because the limitation of very large drop in disdrometer, the third method is not be able to retrieve reasonable DSD at large Z_DR situation. The retrieved three dimensional DSDs in different stages of thunderstorm reveal very bounty microphysical information. The relatively smaller concentration and drop diameter were found in the initiation stage, while the deep convection was developed one to two order of magnitude of concentration were found through the convective core. More cases studies and much longer statistics and validation should be conducted in the near future.