透過您的圖書館登入
IP:18.216.32.116
  • 學位論文

利用WRF 3DVAR同化雷達徑向風對2011年南瑪都颱風模擬之影響

Effects of radar radial wind data assimilation using WRF 3DVAR on the typhoon simulation: Case of typhoon Nanmadol (2011)

指導教授 : 林沛練
若您是本文的作者,可授權文章由華藝線上圖書館中協助推廣。

摘要


台灣地理位置位於熱帶氣旋活躍的西北太平洋地區,為颱風路徑之要衝,每年颱風季中,約有30個颱風生成於西北太平洋地區,當侵台颱風接近台灣時,移動路徑與環流因複雜的地形產生很大的變化,使得台灣地區對颱風的風雨與路徑預報更為困難。   在天氣系統的數值模擬過程中,模式的預報能力常受限於初始場的準確性,全球模式提供的初始場空間解析度不夠高,無法完整描述出中小尺度天氣系統真實大氣的狀態。雷達觀測具有高時空解析度的優點,常使用於劇烈天氣的監控與觀測,能夠觀測颱風環流結構的詳細特徵。本研究挑選2011年8月28日至8月29日侵襲台灣的南瑪都颱風,將雷達高時解析度的徑向風資料同化到模式中,以改善初始場,探討雷達徑向風資料同化對颱風路徑與風雨分布模擬之影響。   本研究使用WRF模式(V3.2.1)和3D-VAR同化系統(WRFDA V3.2.1),使用兩層巢狀網格,水平網格間距為 15和5 公里,網格數為 301 x 253和 241 x 241,垂直分層 35 層。初始時間為2011年8月28日0000 UTC;而WRF 3DVAR cycle使用中央氣象局墾丁雷達(RCKT),同化時間為2011年8月28日0600、0900、1200、1500和1800 UTC,背景場誤差採用NCEP全球模式所計算出的統計值(CV3),並測試兩組水平影響半徑:R1=0.06和R2=0.12,分別進行30、24、21、18、15和12小時模擬,探討不同同化次數和調整其水平影響半徑對數值模擬的影響。   模擬結果在颱風結構上,雷達資料同化會改善近颱風中心的雨帶分佈,並隨著同化循環次數越多,其修正效果越大;而較大的影響半徑 R2的強對流雨帶分布也比 R1 更接近觀測。在颱風移動路徑上,R2平均路徑誤差小於 R1,且隨著同化次數越多,路徑誤差也明顯的下降,證明較大的影響半徑 R2 做多次同化循環明顯改善颱風路徑過快且偏離最佳路徑的情形。在降雨模擬結果中,經過多次同化循環可以改善降雨分佈不正確且高估的情形,而較大影響半徑 R2的強降水分佈位置比R1準確;而雨量驗證上,R1 的 ETS得分在偏弱降水有較高的得分,但在強降雨的模擬上,R2 有較高的準確度。雖然累積雨量還是有高估的情形,但若使用較大的影響半徑配合多次同化循環,會明顯修正雨量的高估和強降水區的分布位置。   經由不同組合的實驗發現:(1)相較於一次雷達資料同化,較多的同化循環次數對颱風模擬的改善效果好;(2)較大的影響半徑(R2=0.12)模擬結果也比較小的影響半徑(R1=0.06)佳。因此以較多的同化循環再加上較大的影響半徑組合實驗能有效改善數值模擬結果,雖然降水仍有高估、颱風移動速度偏快的現象,但相較於未同化實驗的模擬結果,降水分布和颱風路徑有更接近觀測的趨勢;雖然同化效果仍然有限,但就颱風整體降雨、結構和移動路徑而言,同化雷達徑向風資料確實有效的改善模擬結果的準確度。

並列摘要


Taiwan is located in the northeastern pacific region where tropical cyclone is very active and thus a major path for typhoons. Every typhoon season, approximately 30 typhoon forms in the northeastern pacific region. When the typhoon approaches Taiwan, the track and circulation of the typhoon changes dramatically due to the complex geographic features of the island, so it’s difficult to predict the typhoon’s track and rainfall.   Compared with conventional data, radar observations have an advantage of high spatial and temporal resolutions, and Doppler radars are capable of capturing detailed characteristics of flow fields, including typhoon circulation. The observation of radar network in Taiwan was used to investigate the impact of radar data assimilation for typhoon simulation. The case of Typhoon Nanmadol(2011)was chosen for this study, it struck Taiwan from 28 Aug 2011 to 29 Aug 2011. The purpose of this study was to adjust the initial field of the numerical model to improve the short-term typhoon predictions near Taiwan by using Doppler radar radial wind data assimilation. The typhoon track, structure, and precipitation were also inspected to clarify the effect of radar data assimilation simulated when typhoon approaching Taiwan.   In this study, Weather Research and Forecasting(WRF V3.2.1) model and 3D-VAR method of WRF Data Assimilation system(WRFDA V3.2.1)were used in doppler radar data assimilation.. The horizontal grid resolution of nested domains is 15 and 5 km, respectively, horizontal grid points were 301×253 and 241×241, model vertical layers extended from the surface up to 50 hPa with 35 levels. Initial conditions at 0000 UTC 28 Aug, and the WRF 3DVAR cycling from RCKT of CWB at 0600、0900、1200、1500 and 1800 UTC 28 Aug. The 3DVAR influence factor of horizontal scale for radar data assimilation was set R1 = 0.06 and R2 = 0.12.   After a series of experiments, we obtained some conclusions :(1) Compared with the one-time radar data assimilation, cycling of the multiple-time radar data assimilation has more positive impact on typhoon simulation, (2)Larger influence factor of horizontal scale (R2 = 0.12) also has more positive impact than smaller influence factor of horizontal scale(R1 = 0.06)for typhoon simulation in Taiwan. More assimilation cycle collocate with a larger influence factor of horizontal scale(R2 = 0.12), had improved the accuracy of the numerical simulation. Compared with the non-assimilate experiment, the precipitation pattern and typhoon track in the assimilation experiment were closer to the observations.

並列關鍵字

Typhoon Nanmadol 3DVAR

參考文獻


林欣弘,2011:雷達徑向風資料同化對降雨系統模擬之影響研究。國立中央大學,大氣物理研究所,博士論文。
葉天降、吳石吉、謝信良, 1998:台灣附近颱風路徑預報校驗與統計方 法之應用。大氣科學,第26 期3 號,227-248。
Anthes, R. A., 1983: Regional models of the atmosphere in middle latitudes. Mon. Wea.Rev., 111, 1306–1335.
Barker, D. M., W. Huang, Y.-R. Guo, and A. Bourgeois, 2003: A three-dimensional variational (3DVAR) data assimilation system for use with MM5. NCAR Tech. Note, 68 pp.
Barker, D. M., W. Huang, Y.-R. Guo, A. Bourgeois, and Q. Xiao, 2004: A three-dimensional variational data assimilation system for MM5: Implementation and initial results. Mon. Wea. Rev. , 132, 897–914.

延伸閱讀