雷達資料同化於多重尺度天氣系統(梅雨)的強降雨預報影響：SoWMEX IOP#8 個案研究;Impact of the radar data assimilation on heavy rainfall prediction associated with a multi-scale weather (Meiyu) system: a case study of SoWMEX IOP#8

NCUIR > College of Earth Sciences > Department of Atmospheric Sciences and Graduate Institute of Atmospheric Physics > Department of Earth Sciences > Item 987654321/73866

Please use this identifier to cite or link to this item: http://ir.lib.ncu.edu.tw/handle/987654321/73866

Title:	雷達資料同化於多重尺度天氣系統(梅雨)的強降雨預報影響：SoWMEX IOP#8 個案研究;Impact of the radar data assimilation on heavy rainfall prediction associated with a multi-scale weather (Meiyu) system: a case study of SoWMEX IOP#8
Authors:	鄭翔文;Cheng, Hsiang-Wen
Contributors:	大氣科學學系
Keywords:	WRF-LETKF雷達資料同化系統;資料同化;定量降雨即時預報;WRF-LETKF Radar Assimilation System;Data assimilation;quantitative precipitation nowcasting
Date:	2017-08-23
Issue Date:	2017-10-27 12:27:58 (UTC+8)
Publisher:	國立中央大學
Abstract:	本研究使用Tsai et al. (2014) 開發之雷達資料同化系統(WRF-LETKF Radar Assimilation System) 同化七股及墾丁雷達資料，並提出有用的同化策略，以改善IOP#8期間的6月16日的降雨表現，探討此個案在臺灣西南至南部沿海地區所造成的持續性降雨能否藉由同化過中尺度觀測資訊的分析系集搭配WLRAS來改善梅雨個案的短期預報降水能力。梅雨季的強降雨事件富含多重尺度交互作用與地形效應，造成在定量降水預報上的困難，因此本研究使用同化過中尺度資料之分析系集做為初始場，再利用對流尺度的雷達資料同化改善對流尺度的預報，藉由結合中尺度資料同化與對流尺度資料同化來改善梅雨季的降水預報。本研究採用三種不同的初始系集進行測試，分別為：1) Yang et al. (2014) 同化過傳統氣象觀測資料、衛星觀測資料之分析系集，2) 同上設定，但沒有GPS-RO觀測資料之分析系集，及3) 將NECP GFS Final Analysis (FNL 1^°×1^° 資料) 以3DVAR的背景誤差協方差加入隨機擾動而成的系集；實驗結果說明當提供同化過中尺度觀測資訊之環境場且含有流場相依的誤差結構作為初始場有利於掌握此個案之降水分布。本研究亦測試WLRAS的同化策略，我們分別實驗同化徑向風更新模式風場、同化回波更新水氣相關變數、同化徑向風與回波更新所有模式預報變數。結果指出，僅修正風場可改善降雨的時間分布、但降雨強度與空間分布則仍有待改善；僅修正水氣相關變數則可改善降雨強度與空間分布，但水氣相關變數缺乏風場支持，在預報前期形成大量降雨；同時修正風場與水氣相關變數則可改善降雨的時間分布、空間分布及降雨強度，是為較佳的資料同化策略。此外，雷達觀測風場有無法觀測切向風的限制，當使用觀測之徑向風同化切向之風場時，可能因為取樣誤差等誤差導致風場有錯誤的調整，降低分析品質。為減少上述情形，本實驗在雷達觀測範圍重複區採用特殊的觀測品質控管策略，以提升風場品質。;This study applies the WRF-LETKF Radar Assimilation System (WLRAS; Tsai et al. 2014), which couples the local ensemble transform Kalman filter with Weather Research and Forecasting, to assimilate two Doppler radars in Taiwan. The initial condition is the initial ensembles from Yang et al. (2014) which assimilate GTS and satellite data or only assimilate GTS data and AMV data or NCEP FNL data as initial state to improve short-term quantitative precipitation forecast. The importance of the initial fields to quantitative precipitation nowcasting are evaluated based on 2008 SoWMEX IOP#8 (2008/06/16). This study explores useful radar data assimilation strategies for improving the heavy precipitation prediction during Meiyu seasons in Taiwan, which are challenges due to complex terrain and multi-scale interactions. Results show that providing a good environment and perturbation for WLRAS can improve the heavy precipitation prediction during Meiyu seasons in Taiwan. Results also indicate that assimilating radial wind can improve wind field and assimilating reflectivity can improve QPF but the temporal evolution is wrong. It is found that assimilating radial wind and reflectivity to update all model variables is a better assimilation strategy then only updating wind or water-related variables. Furthermore, due to the limitation from no tangential wind observed by radars, we remove the radial wind in the two radar observation coverage when updating u/v. The improved QC strategies can improve wind field, further improve QPF.
Appears in Collections:	[Department of Atmospheric Sciences and Graduate Institute of Atmospheric Physics ] Department of Earth Sciences

Files in This Item:

File	Description	Size	Format
index.html		0Kb	HTML	655	View/Open

社群 sharing

Loading...